GRU_Classification.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KbFrd57eiBvD"
      },
      "source": [
        "<center><h1>Projet Robotique</h1></center>\n",
        "<center><h2>Classification via GRU</h2></center>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "855UIY6UiJeb"
      },
      "source": [
        "# Importation des librairies"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QTIsZfAL_sOz",
        "outputId": "32b5d753-3d32-498a-ac55-82c0c6c52517"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Collecting bayesian-optimization\n",
            "  Downloading bayesian_optimization-1.4.3-py3-none-any.whl (18 kB)\n",
            "Requirement already satisfied: numpy>=1.9.0 in /usr/local/lib/python3.10/dist-packages (from bayesian-optimization) (1.23.5)\n",
            "Requirement already satisfied: scipy>=1.0.0 in /usr/local/lib/python3.10/dist-packages (from bayesian-optimization) (1.11.4)\n",
            "Requirement already satisfied: scikit-learn>=0.18.0 in /usr/local/lib/python3.10/dist-packages (from bayesian-optimization) (1.2.2)\n",
            "Collecting colorama>=0.4.6 (from bayesian-optimization)\n",
            "  Downloading colorama-0.4.6-py2.py3-none-any.whl (25 kB)\n",
            "Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn>=0.18.0->bayesian-optimization) (1.3.2)\n",
            "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn>=0.18.0->bayesian-optimization) (3.2.0)\n",
            "Installing collected packages: colorama, bayesian-optimization\n",
            "Successfully installed bayesian-optimization-1.4.3 colorama-0.4.6\n"
          ]
        }
      ],
      "source": [
        "!pip install bayesian-optimization"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "uG4PVG23MdLq"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "from google.colab import drive\n",
        "\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "from matplotlib import pyplot as plt\n",
        "import seaborn as sns\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.decomposition import PCA\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.metrics import classification_report\n",
        "\n",
        "from bayes_opt import BayesianOptimization\n",
        "\n",
        "import tensorflow as tf\n",
        "from keras.utils import to_categorical\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense, GRU, Dropout, Flatten\n",
        "from keras.optimizers import Adam\n",
        "from keras.api._v2.keras.callbacks import ModelCheckpoint, EarlyStopping\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "# disable tensorflow warnings\n",
        "os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "t-3EzW1siLrc"
      },
      "source": [
        "# Configaration GPU"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XPJKH4ywNRdq",
        "outputId": "6457ea80-f3f1-450a-d36b-39a8a15df401"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Found GPU at:/device:GPU:0\n"
          ]
        }
      ],
      "source": [
        "device_name = tf.test.gpu_device_name()\n",
        "if len(device_name) > 0:\n",
        "    print(\"Found GPU at:{}\".format(device_name))\n",
        "else:\n",
        "    device_name=\"/device:CPU:0\"\n",
        "    print(\"No GPU\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pgG3WW5yiRRr"
      },
      "source": [
        "# Importation du dataset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pdSWSsLtNcMY",
        "outputId": "cd216b60-28de-49a3-d1b0-e3c73d894d9d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "drive.mount('/content/drive')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 461
        },
        "id": "bJXFWLPWNwSs",
        "outputId": "7692ecc3-a740-4796-9352-d76227773adf"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-67240398-9fbc-47a4-be6b-7cf4260c609c\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>MILLISEC</th>\n",
              "      <th>Accelerometer RKN^ accX</th>\n",
              "      <th>Accelerometer RKN^ accY</th>\n",
              "      <th>Accelerometer RKN^ accZ</th>\n",
              "      <th>Accelerometer HIP accX</th>\n",
              "      <th>Accelerometer HIP accY</th>\n",
              "      <th>Accelerometer HIP accZ</th>\n",
              "      <th>Accelerometer LUA^ accX</th>\n",
              "      <th>Accelerometer LUA^ accY</th>\n",
              "      <th>Accelerometer LUA^ accZ</th>\n",
              "      <th>...</th>\n",
              "      <th>LOCATION TAG4 X</th>\n",
              "      <th>LOCATION TAG4 Y</th>\n",
              "      <th>LOCATION TAG4 Z</th>\n",
              "      <th>Locomotion</th>\n",
              "      <th>HL_Activity</th>\n",
              "      <th>LL_Left_Arm</th>\n",
              "      <th>LL_Left_Arm_Object</th>\n",
              "      <th>LL_Right_Arm</th>\n",
              "      <th>LL_Right_Arm_Object</th>\n",
              "      <th>ML_Both_Arms</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>-1.585284</td>\n",
              "      <td>0.146747</td>\n",
              "      <td>0.337532</td>\n",
              "      <td>-1.410530</td>\n",
              "      <td>0.693260</td>\n",
              "      <td>0.434434</td>\n",
              "      <td>0.074936</td>\n",
              "      <td>-0.199117</td>\n",
              "      <td>0.621667</td>\n",
              "      <td>-0.455638</td>\n",
              "      <td>...</td>\n",
              "      <td>0.147947</td>\n",
              "      <td>0.811792</td>\n",
              "      <td>0.186199</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>-1.585185</td>\n",
              "      <td>0.263802</td>\n",
              "      <td>0.342378</td>\n",
              "      <td>-1.626921</td>\n",
              "      <td>0.626338</td>\n",
              "      <td>0.487110</td>\n",
              "      <td>0.200034</td>\n",
              "      <td>-0.108635</td>\n",
              "      <td>0.598101</td>\n",
              "      <td>-0.542260</td>\n",
              "      <td>...</td>\n",
              "      <td>0.147947</td>\n",
              "      <td>0.812691</td>\n",
              "      <td>0.178270</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>-1.585083</td>\n",
              "      <td>0.194202</td>\n",
              "      <td>0.371457</td>\n",
              "      <td>-1.735116</td>\n",
              "      <td>0.470187</td>\n",
              "      <td>0.527630</td>\n",
              "      <td>0.179184</td>\n",
              "      <td>-0.099275</td>\n",
              "      <td>0.709202</td>\n",
              "      <td>-0.572130</td>\n",
              "      <td>...</td>\n",
              "      <td>0.147947</td>\n",
              "      <td>0.814490</td>\n",
              "      <td>0.172322</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>-1.584985</td>\n",
              "      <td>0.058164</td>\n",
              "      <td>0.153369</td>\n",
              "      <td>-1.616313</td>\n",
              "      <td>0.618902</td>\n",
              "      <td>0.523578</td>\n",
              "      <td>0.210459</td>\n",
              "      <td>-0.152316</td>\n",
              "      <td>0.729402</td>\n",
              "      <td>-0.593039</td>\n",
              "      <td>...</td>\n",
              "      <td>0.147947</td>\n",
              "      <td>0.816289</td>\n",
              "      <td>0.172322</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>-1.584886</td>\n",
              "      <td>0.247984</td>\n",
              "      <td>0.290684</td>\n",
              "      <td>-1.705415</td>\n",
              "      <td>0.570570</td>\n",
              "      <td>0.535734</td>\n",
              "      <td>0.161809</td>\n",
              "      <td>-0.189757</td>\n",
              "      <td>0.709202</td>\n",
              "      <td>-0.634857</td>\n",
              "      <td>...</td>\n",
              "      <td>0.149055</td>\n",
              "      <td>0.818988</td>\n",
              "      <td>0.176287</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606043</th>\n",
              "      <td>1.407215</td>\n",
              "      <td>0.640278</td>\n",
              "      <td>0.298761</td>\n",
              "      <td>-0.141884</td>\n",
              "      <td>0.358651</td>\n",
              "      <td>0.102168</td>\n",
              "      <td>0.787298</td>\n",
              "      <td>0.418657</td>\n",
              "      <td>0.540867</td>\n",
              "      <td>-0.067329</td>\n",
              "      <td>...</td>\n",
              "      <td>0.129671</td>\n",
              "      <td>0.139901</td>\n",
              "      <td>0.208006</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606044</th>\n",
              "      <td>1.407317</td>\n",
              "      <td>0.437804</td>\n",
              "      <td>0.297145</td>\n",
              "      <td>-0.020960</td>\n",
              "      <td>0.291729</td>\n",
              "      <td>0.102168</td>\n",
              "      <td>0.776873</td>\n",
              "      <td>0.452977</td>\n",
              "      <td>0.483633</td>\n",
              "      <td>-0.195770</td>\n",
              "      <td>...</td>\n",
              "      <td>0.129117</td>\n",
              "      <td>0.138102</td>\n",
              "      <td>0.206023</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606045</th>\n",
              "      <td>1.407416</td>\n",
              "      <td>0.678242</td>\n",
              "      <td>0.240604</td>\n",
              "      <td>-0.001866</td>\n",
              "      <td>0.143014</td>\n",
              "      <td>0.110272</td>\n",
              "      <td>0.915871</td>\n",
              "      <td>0.393696</td>\n",
              "      <td>0.503833</td>\n",
              "      <td>-0.195770</td>\n",
              "      <td>...</td>\n",
              "      <td>0.129117</td>\n",
              "      <td>0.136303</td>\n",
              "      <td>0.202059</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606046</th>\n",
              "      <td>1.407514</td>\n",
              "      <td>1.155955</td>\n",
              "      <td>0.159831</td>\n",
              "      <td>-0.046417</td>\n",
              "      <td>0.098399</td>\n",
              "      <td>0.256145</td>\n",
              "      <td>0.780348</td>\n",
              "      <td>0.362495</td>\n",
              "      <td>0.577900</td>\n",
              "      <td>-0.276418</td>\n",
              "      <td>...</td>\n",
              "      <td>0.129671</td>\n",
              "      <td>0.134504</td>\n",
              "      <td>0.200076</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606047</th>\n",
              "      <td>1.407616</td>\n",
              "      <td>0.323912</td>\n",
              "      <td>0.445768</td>\n",
              "      <td>-0.018838</td>\n",
              "      <td>0.131860</td>\n",
              "      <td>0.199417</td>\n",
              "      <td>0.915871</td>\n",
              "      <td>0.318815</td>\n",
              "      <td>0.651968</td>\n",
              "      <td>-0.106160</td>\n",
              "      <td>...</td>\n",
              "      <td>0.130224</td>\n",
              "      <td>0.133605</td>\n",
              "      <td>0.198094</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>606048 rows × 250 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-67240398-9fbc-47a4-be6b-7cf4260c609c')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-67240398-9fbc-47a4-be6b-7cf4260c609c button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-67240398-9fbc-47a4-be6b-7cf4260c609c');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-bc05ab8a-68bf-4c20-91d9-04126487c70e\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-bc05ab8a-68bf-4c20-91d9-04126487c70e')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-bc05ab8a-68bf-4c20-91d9-04126487c70e button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "text/plain": [
              "        MILLISEC  Accelerometer RKN^ accX  Accelerometer RKN^ accY  \\\n",
              "0      -1.585284                 0.146747                 0.337532   \n",
              "1      -1.585185                 0.263802                 0.342378   \n",
              "2      -1.585083                 0.194202                 0.371457   \n",
              "3      -1.584985                 0.058164                 0.153369   \n",
              "4      -1.584886                 0.247984                 0.290684   \n",
              "...          ...                      ...                      ...   \n",
              "606043  1.407215                 0.640278                 0.298761   \n",
              "606044  1.407317                 0.437804                 0.297145   \n",
              "606045  1.407416                 0.678242                 0.240604   \n",
              "606046  1.407514                 1.155955                 0.159831   \n",
              "606047  1.407616                 0.323912                 0.445768   \n",
              "\n",
              "        Accelerometer RKN^ accZ  Accelerometer HIP accX  \\\n",
              "0                     -1.410530                0.693260   \n",
              "1                     -1.626921                0.626338   \n",
              "2                     -1.735116                0.470187   \n",
              "3                     -1.616313                0.618902   \n",
              "4                     -1.705415                0.570570   \n",
              "...                         ...                     ...   \n",
              "606043                -0.141884                0.358651   \n",
              "606044                -0.020960                0.291729   \n",
              "606045                -0.001866                0.143014   \n",
              "606046                -0.046417                0.098399   \n",
              "606047                -0.018838                0.131860   \n",
              "\n",
              "        Accelerometer HIP accY  Accelerometer HIP accZ  \\\n",
              "0                     0.434434                0.074936   \n",
              "1                     0.487110                0.200034   \n",
              "2                     0.527630                0.179184   \n",
              "3                     0.523578                0.210459   \n",
              "4                     0.535734                0.161809   \n",
              "...                        ...                     ...   \n",
              "606043                0.102168                0.787298   \n",
              "606044                0.102168                0.776873   \n",
              "606045                0.110272                0.915871   \n",
              "606046                0.256145                0.780348   \n",
              "606047                0.199417                0.915871   \n",
              "\n",
              "        Accelerometer LUA^ accX  Accelerometer LUA^ accY  \\\n",
              "0                     -0.199117                 0.621667   \n",
              "1                     -0.108635                 0.598101   \n",
              "2                     -0.099275                 0.709202   \n",
              "3                     -0.152316                 0.729402   \n",
              "4                     -0.189757                 0.709202   \n",
              "...                         ...                      ...   \n",
              "606043                 0.418657                 0.540867   \n",
              "606044                 0.452977                 0.483633   \n",
              "606045                 0.393696                 0.503833   \n",
              "606046                 0.362495                 0.577900   \n",
              "606047                 0.318815                 0.651968   \n",
              "\n",
              "        Accelerometer LUA^ accZ  ...  LOCATION TAG4 X  LOCATION TAG4 Y  \\\n",
              "0                     -0.455638  ...         0.147947         0.811792   \n",
              "1                     -0.542260  ...         0.147947         0.812691   \n",
              "2                     -0.572130  ...         0.147947         0.814490   \n",
              "3                     -0.593039  ...         0.147947         0.816289   \n",
              "4                     -0.634857  ...         0.149055         0.818988   \n",
              "...                         ...  ...              ...              ...   \n",
              "606043                -0.067329  ...         0.129671         0.139901   \n",
              "606044                -0.195770  ...         0.129117         0.138102   \n",
              "606045                -0.195770  ...         0.129117         0.136303   \n",
              "606046                -0.276418  ...         0.129671         0.134504   \n",
              "606047                -0.106160  ...         0.130224         0.133605   \n",
              "\n",
              "        LOCATION TAG4 Z  Locomotion  HL_Activity  LL_Left_Arm  \\\n",
              "0              0.186199           0            0            0   \n",
              "1              0.178270           0            0            0   \n",
              "2              0.172322           0            0            0   \n",
              "3              0.172322           0            0            0   \n",
              "4              0.176287           0            0            0   \n",
              "...                 ...         ...          ...          ...   \n",
              "606043         0.208006           0            0            0   \n",
              "606044         0.206023           0            0            0   \n",
              "606045         0.202059           0            0            0   \n",
              "606046         0.200076           0            0            0   \n",
              "606047         0.198094           0            0            0   \n",
              "\n",
              "        LL_Left_Arm_Object  LL_Right_Arm  LL_Right_Arm_Object  ML_Both_Arms  \n",
              "0                        0             0                    0             0  \n",
              "1                        0             0                    0             0  \n",
              "2                        0             0                    0             0  \n",
              "3                        0             0                    0             0  \n",
              "4                        0             0                    0             0  \n",
              "...                    ...           ...                  ...           ...  \n",
              "606043                   0             0                    0             0  \n",
              "606044                   0             0                    0             0  \n",
              "606045                   0             0                    0             0  \n",
              "606046                   0             0                    0             0  \n",
              "606047                   0             0                    0             0  \n",
              "\n",
              "[606048 rows x 250 columns]"
            ]
          },
          "execution_count": 6,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df = pd.read_csv(\"/content/drive/MyDrive/Robotique/preprocessed_data.csv\")\n",
        "\n",
        "df"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4aEEuOpBiY_O"
      },
      "source": [
        "# Fonctions diverses"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PFCBgbyWikoe"
      },
      "source": [
        "# Fonctions de feature selection"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "SWl-S_hWkaj3"
      },
      "outputs": [],
      "source": [
        "def dim_reduction(df, n):\n",
        "    \"\"\"\n",
        "    Perform dimensionality reduction using PCA.\n",
        "\n",
        "    Args:\n",
        "    df (pandas.DataFrame): Input DataFrame.\n",
        "    n (int): Number of components for dimensionality reduction.\n",
        "\n",
        "    Returns:\n",
        "    pandas.DataFrame: DataFrame with reduced dimensions.\n",
        "    \"\"\"\n",
        "    pca = PCA(n_components=n)\n",
        "    pca.fit(df)\n",
        "    reducted_df = pca.fit_transform(df)\n",
        "\n",
        "    # Create column names for the reduced dimensions\n",
        "    columns_name = [f'column{i}' for i in range(1, pca.n_components_ + 1)]\n",
        "\n",
        "    return pd.DataFrame(reducted_df, columns=columns_name)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "yAXrZ6fjXNW0"
      },
      "outputs": [],
      "source": [
        "def correlation(dataset, threshold):\n",
        "    \"\"\"\n",
        "    Remove columns with high correlation from the input DataFrame.\n",
        "\n",
        "    Args:\n",
        "    dataset (pandas.DataFrame): Input DataFrame.\n",
        "    threshold (float): Correlation threshold for column removal.\n",
        "\n",
        "    Returns:\n",
        "    pandas.DataFrame: DataFrame with correlated columns removed.\n",
        "    \"\"\"\n",
        "    df = dataset.copy()\n",
        "    col_corr = set()  # For storing unique columns with high correlation\n",
        "    corr_matrix = dataset.corr()\n",
        "\n",
        "    # Iterate through the correlation matrix\n",
        "    for i in range(len(corr_matrix.columns)):\n",
        "        for j in range(i):\n",
        "            # Check if the absolute correlation value is greater than the threshold\n",
        "            if abs(corr_matrix.iloc[i, j]) > threshold:\n",
        "                colname = corr_matrix.columns[i]\n",
        "                col_corr.add(colname)\n",
        "\n",
        "    # Drop columns with high correlation\n",
        "    df.drop(col_corr, axis=1, inplace=True)\n",
        "\n",
        "    return df"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y1RAFqUFipHU"
      },
      "source": [
        "## Fonction de segmentation du dataset  "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ZnrUvuO-OAyJ"
      },
      "outputs": [],
      "source": [
        "def data_segmentation(data_collection, window_size, label):\n",
        "    \"\"\"\n",
        "    Segment the data and create a dataset with a choosen labels.\n",
        "\n",
        "    Args:\n",
        "    data_collection (pandas.DataFrame): Input DataFrame.\n",
        "    window_size (int): Size of the segmentation window.\n",
        "    label (str): Name of the column to be used as the target label.\n",
        "\n",
        "    Returns:\n",
        "    numpy.ndarray: Segmented input features (X).\n",
        "    numpy.ndarray: Corresponding labels (y).\n",
        "    \"\"\"\n",
        "    loco_i = data_collection.columns.get_loc(label)\n",
        "\n",
        "    # Convert the DataFrame to a numpy array\n",
        "    data = data_collection.to_numpy()\n",
        "\n",
        "    # Segment the data\n",
        "    n = len(data)\n",
        "    X = []\n",
        "    y = []\n",
        "    start = 0\n",
        "    end = 0\n",
        "\n",
        "    while start + window_size - 1 < n:\n",
        "        end = start + window_size - 1\n",
        "\n",
        "        # Check if the frame contains the same activity and is from the same file\n",
        "        if data[start][loco_i] == data[end][loco_i] and data[start][-1] == data[end][-1]:\n",
        "            X.append(data[start:(end+1), :])\n",
        "            y.append(data[start][loco_i])\n",
        "            start += window_size // 2  # 50% overlap\n",
        "        else:\n",
        "            # If the frame contains different activities or is from different objects, find the next start point\n",
        "            while start + window_size - 1 < n:\n",
        "                if data[start][loco_i] != data[start+1][loco_i]:\n",
        "                    break\n",
        "                start += 1\n",
        "            start += 1\n",
        "\n",
        "    return np.asarray(X), np.asarray(y)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AfNaBl_yiwY6"
      },
      "source": [
        "## Fonction d'entrainement du modèle"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "tEX4GgSMOkV5"
      },
      "outputs": [],
      "source": [
        "def gru_model(X_train, y_train, X_val, y_val, nbn_g, nbn_d, lr, bs, dp, v):\n",
        "    nbn_g = int(nbn_g)\n",
        "    nbn_d = int(nbn_d)\n",
        "    bs = int(bs)\n",
        "    with tf.device(device_name):\n",
        "\n",
        "      model = Sequential()\n",
        "\n",
        "      model.add(GRU(units=nbn_g, activation='tanh', return_sequences=True, input_shape=X_train[0].shape))\n",
        "\n",
        "      model.add(GRU(units=nbn_g//2, activation='tanh', return_sequences=True, input_shape=X_train[0].shape))\n",
        "\n",
        "      model.add(GRU(units=nbn_g//4, activation='tanh', return_sequences=True, input_shape=X_train[0].shape))\n",
        "\n",
        "      model.add(GRU(units=nbn_g//8, activation='tanh'))\n",
        "\n",
        "      model.add(Flatten())\n",
        "      model.add(Dense(nbn_d, activation='elu'))\n",
        "      model.add(Dropout(dp))\n",
        "      model.add(Dense(nbn_g//2, activation='elu'))\n",
        "      model.add(Dropout(dp))\n",
        "      model.add(Dense(5, activation='softmax'))\n",
        "\n",
        "\n",
        "      model.compile(optimizer = Adam(learning_rate=lr),\n",
        "                    loss = 'categorical_crossentropy',\n",
        "                    metrics =[tf.keras.metrics.TruePositives(name='tp'),\n",
        "                          tf.keras.metrics.FalsePositives(name='fp'),\n",
        "                          tf.keras.metrics.TrueNegatives(name='tn'),\n",
        "                          tf.keras.metrics.FalseNegatives(name='fn'),\n",
        "                          'accuracy',\n",
        "                          tf.keras.metrics.Precision(name='precision'),\n",
        "                          tf.keras.metrics.Recall(name='recall'),\n",
        "                          tf.keras.metrics.AUC(name='auc')])\n",
        "\n",
        "      checkpointer = ModelCheckpoint(filepath='test.keras', monitor='accuracy', verbose=0, save_best_only=True, save_weights_only=False, mode='min', save_freq=1)\n",
        "      early = EarlyStopping(monitor='val_loss', min_delta=0, patience=5, mode='auto')\n",
        "\n",
        "      history = model.fit(X_train, y_train,\n",
        "                              batch_size=bs, epochs=100,\n",
        "                              verbose=v,\n",
        "                              callbacks = [checkpointer, early],\n",
        "                              validation_data=(X_val, y_val))\n",
        "\n",
        "      return history, model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HdkiJ2GFi3dB"
      },
      "source": [
        "## Fonction de la pipeline à utiliser"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "TsI0ddB4YW4h"
      },
      "outputs": [],
      "source": [
        "def pipeline(method,tresh,nbn_g, nbn_d, lr, bs, dp):\n",
        "  \"\"\"\n",
        "    Perform a GRU model pipeline with dimensionality reduction or correlation removal.\n",
        "\n",
        "    Args:\n",
        "    method (int): Method for data preprocessing. 1: Dimensionality reduction, 2: Correlation removal.\n",
        "    threshold (float): Threshold for dimensionality reduction or correlation removal.\n",
        "    nbn_g (int): Number of GRU units.\n",
        "    nbn_d (int): Number of dense units.\n",
        "    lr (float): Learning rate.\n",
        "    bs (int): Batch size.\n",
        "    dp (float): Dropout rate.\n",
        "\n",
        "    Returns:\n",
        "    float: Accuracy score of the GRU model.\n",
        "    \"\"\"\n",
        "\n",
        "  if method <= 1:\n",
        "    dataset = dim_reduction(df[df.columns[:-7]], tresh)\n",
        "  elif method <= 2:\n",
        "    dataset = correlation(df[df.columns[:-7]], tresh)\n",
        "\n",
        "  labels = df[df.columns[-7:]]\n",
        "  dataset = pd.concat([dataset, labels], axis=1)\n",
        "\n",
        "  X, y = data_segmentation(data_collection=dataset, window_size=25, label='HL_Activity')\n",
        "\n",
        "  y = to_categorical(y)\n",
        "\n",
        "  # split the data into train, validation and test sets using sklearn\n",
        "  X_train, X_test, y_train, y_test = train_test_split(X, y, shuffle=True, test_size=0.33)\n",
        "  X_test, X_val, y_test, y_val = train_test_split(X_test, y_test, shuffle=True, test_size=0.5)\n",
        "\n",
        "  history, model = gru_model(X_train, y_train, X_val, y_val, nbn_g, nbn_d, lr, bs, dp, 0)\n",
        "  acc = history.history['val_accuracy'][-1]\n",
        "\n",
        "  return acc"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "W4xgFw_WjSI5"
      },
      "source": [
        "# Affinage de la pipeline via l'optimisation Bayesienne"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xPxF_3gaVJdL",
        "outputId": "3afaa595-c7ad-4787-e11e-c738333778fd"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "|   iter    |  target   |    bs     |    dp     |    lr     |  method   |   nbn_d   |   nbn_l   |   tresh   |\n",
            "-------------------------------------------------------------------------------------------------------------\n",
            "| \u001b[0m1        \u001b[0m | \u001b[0m0.7474   \u001b[0m | \u001b[0m232.2    \u001b[0m | \u001b[0m0.3881   \u001b[0m | \u001b[0m2.144e-06\u001b[0m | \u001b[0m0.6047   \u001b[0m | \u001b[0m368.7    \u001b[0m | \u001b[0m326.9    \u001b[0m | \u001b[0m0.2658   \u001b[0m |\n",
            "| \u001b[95m2        \u001b[0m | \u001b[95m0.9794   \u001b[0m | \u001b[95m197.9    \u001b[0m | \u001b[95m0.2587   \u001b[0m | \u001b[95m0.005389 \u001b[0m | \u001b[95m0.8384   \u001b[0m | \u001b[95m782.2    \u001b[0m | \u001b[95m413.0    \u001b[0m | \u001b[95m0.8815   \u001b[0m |\n",
            "| \u001b[95m3        \u001b[0m | \u001b[95m0.9985   \u001b[0m | \u001b[95m45.15    \u001b[0m | \u001b[95m0.3682   \u001b[0m | \u001b[95m0.004174 \u001b[0m | \u001b[95m1.117    \u001b[0m | \u001b[95m363.8    \u001b[0m | \u001b[95m408.1    \u001b[0m | \u001b[95m0.8127   \u001b[0m |\n",
            "| \u001b[95m4        \u001b[0m | \u001b[95m1.0      \u001b[0m | \u001b[95m496.8    \u001b[0m | \u001b[95m0.2254   \u001b[0m | \u001b[95m0.006924 \u001b[0m | \u001b[95m1.753    \u001b[0m | \u001b[95m943.1    \u001b[0m | \u001b[95m321.3    \u001b[0m | \u001b[95m0.1348   \u001b[0m |\n",
            "| \u001b[0m5        \u001b[0m | \u001b[0m0.9927   \u001b[0m | \u001b[0m113.5    \u001b[0m | \u001b[0m0.4513   \u001b[0m | \u001b[0m0.0009844\u001b[0m | \u001b[0m0.8422   \u001b[0m | \u001b[0m991.7    \u001b[0m | \u001b[0m665.5    \u001b[0m | \u001b[0m0.7158   \u001b[0m |\n",
            "| \u001b[0m6        \u001b[0m | \u001b[0m0.975    \u001b[0m | \u001b[0m183.4    \u001b[0m | \u001b[0m0.3746   \u001b[0m | \u001b[0m0.008346 \u001b[0m | \u001b[0m0.03658  \u001b[0m | \u001b[0m832.1    \u001b[0m | \u001b[0m1.015e+03\u001b[0m | \u001b[0m0.7659   \u001b[0m |\n",
            "| \u001b[0m7        \u001b[0m | \u001b[0m0.9985   \u001b[0m | \u001b[0m166.6    \u001b[0m | \u001b[0m0.4157   \u001b[0m | \u001b[0m0.001033 \u001b[0m | \u001b[0m0.8958   \u001b[0m | \u001b[0m953.8    \u001b[0m | \u001b[0m481.5    \u001b[0m | \u001b[0m0.3561   \u001b[0m |\n",
            "| \u001b[0m8        \u001b[0m | \u001b[0m0.9838   \u001b[0m | \u001b[0m94.41    \u001b[0m | \u001b[0m0.1077   \u001b[0m | \u001b[0m0.006789 \u001b[0m | \u001b[0m0.4233   \u001b[0m | \u001b[0m459.9    \u001b[0m | \u001b[0m633.5    \u001b[0m | \u001b[0m0.1475   \u001b[0m |\n",
            "| \u001b[0m9        \u001b[0m | \u001b[0m0.9838   \u001b[0m | \u001b[0m307.6    \u001b[0m | \u001b[0m0.1587   \u001b[0m | \u001b[0m0.005893 \u001b[0m | \u001b[0m1.4      \u001b[0m | \u001b[0m334.6    \u001b[0m | \u001b[0m574.0    \u001b[0m | \u001b[0m0.718    \u001b[0m |\n",
            "| \u001b[0m10       \u001b[0m | \u001b[0m0.6887   \u001b[0m | \u001b[0m230.8    \u001b[0m | \u001b[0m0.12     \u001b[0m | \u001b[0m0.005359 \u001b[0m | \u001b[0m1.328    \u001b[0m | \u001b[0m651.4    \u001b[0m | \u001b[0m981.4    \u001b[0m | \u001b[0m0.622    \u001b[0m |\n",
            "| \u001b[0m11       \u001b[0m | \u001b[0m0.9868   \u001b[0m | \u001b[0m50.88    \u001b[0m | \u001b[0m0.2038   \u001b[0m | \u001b[0m0.007302 \u001b[0m | \u001b[0m1.62     \u001b[0m | \u001b[0m358.3    \u001b[0m | \u001b[0m401.0    \u001b[0m | \u001b[0m0.8277   \u001b[0m |\n",
            "| \u001b[0m12       \u001b[0m | \u001b[0m0.9604   \u001b[0m | \u001b[0m56.57    \u001b[0m | \u001b[0m0.4277   \u001b[0m | \u001b[0m0.008047 \u001b[0m | \u001b[0m0.8622   \u001b[0m | \u001b[0m329.5    \u001b[0m | \u001b[0m558.4    \u001b[0m | \u001b[0m0.656    \u001b[0m |\n",
            "| \u001b[0m13       \u001b[0m | \u001b[0m0.9985   \u001b[0m | \u001b[0m32.0     \u001b[0m | \u001b[0m0.5      \u001b[0m | \u001b[0m0.0008406\u001b[0m | \u001b[0m0.0      \u001b[0m | \u001b[0m515.9    \u001b[0m | \u001b[0m471.2    \u001b[0m | \u001b[0m0.6762   \u001b[0m |\n",
            "| \u001b[0m14       \u001b[0m | \u001b[0m0.7034   \u001b[0m | \u001b[0m364.5    \u001b[0m | \u001b[0m0.1896   \u001b[0m | \u001b[0m1e-06    \u001b[0m | \u001b[0m1.413    \u001b[0m | \u001b[0m929.0    \u001b[0m | \u001b[0m481.2    \u001b[0m | \u001b[0m0.6892   \u001b[0m |\n",
            "| \u001b[0m15       \u001b[0m | \u001b[0m0.9868   \u001b[0m | \u001b[0m74.89    \u001b[0m | \u001b[0m0.2927   \u001b[0m | \u001b[0m0.00359  \u001b[0m | \u001b[0m0.9009   \u001b[0m | \u001b[0m891.3    \u001b[0m | \u001b[0m552.4    \u001b[0m | \u001b[0m0.4511   \u001b[0m |\n",
            "| \u001b[0m16       \u001b[0m | \u001b[0m0.9251   \u001b[0m | \u001b[0m72.07    \u001b[0m | \u001b[0m0.5      \u001b[0m | \u001b[0m0.01     \u001b[0m | \u001b[0m0.7307   \u001b[0m | \u001b[0m897.4    \u001b[0m | \u001b[0m386.5    \u001b[0m | \u001b[0m0.4252   \u001b[0m |\n",
            "| \u001b[0m17       \u001b[0m | \u001b[0m0.9648   \u001b[0m | \u001b[0m62.97    \u001b[0m | \u001b[0m0.4943   \u001b[0m | \u001b[0m0.008146 \u001b[0m | \u001b[0m0.9904   \u001b[0m | \u001b[0m1.019e+03\u001b[0m | \u001b[0m521.9    \u001b[0m | \u001b[0m0.4653   \u001b[0m |\n",
            "| \u001b[0m18       \u001b[0m | \u001b[0m0.8488   \u001b[0m | \u001b[0m111.2    \u001b[0m | \u001b[0m0.2017   \u001b[0m | \u001b[0m1e-06    \u001b[0m | \u001b[0m0.6559   \u001b[0m | \u001b[0m651.5    \u001b[0m | \u001b[0m536.0    \u001b[0m | \u001b[0m0.6497   \u001b[0m |\n",
            "| \u001b[0m19       \u001b[0m | \u001b[0m0.9648   \u001b[0m | \u001b[0m112.6    \u001b[0m | \u001b[0m0.4937   \u001b[0m | \u001b[0m0.009347 \u001b[0m | \u001b[0m1.236    \u001b[0m | \u001b[0m430.8    \u001b[0m | \u001b[0m482.0    \u001b[0m | \u001b[0m0.5217   \u001b[0m |\n",
            "| \u001b[0m20       \u001b[0m | \u001b[0m0.9325   \u001b[0m | \u001b[0m214.3    \u001b[0m | \u001b[0m0.3319   \u001b[0m | \u001b[0m0.01     \u001b[0m | \u001b[0m1.33     \u001b[0m | \u001b[0m361.1    \u001b[0m | \u001b[0m694.7    \u001b[0m | \u001b[0m0.3782   \u001b[0m |\n",
            "=============================================================================================================\n",
            "{'target': 1.0, 'params': {'bs': 496.7655563453108, 'dp': 0.22536967126369714, 'lr': 0.006923533834077472, 'method': 1.7527783045920766, 'nbn_d': 943.0579175709547, 'nbn_l': 321.3139543319894, 'tresh': 0.1347587570772653}}\n"
          ]
        }
      ],
      "source": [
        "# Define parameter bounds for optimization\n",
        "pbounds = {\n",
        "    'method': (0, 2),\n",
        "    'tresh': (0.1, 0.99),\n",
        "    'nbn_l': (256, 1024),\n",
        "    'nbn_d': (256, 1024),\n",
        "    'lr': (1e-6, 1e-2),\n",
        "    'bs': (32, 512),\n",
        "    'dp': (0.1, 0.5)\n",
        "    }\n",
        "\n",
        "# Create BayesianOptimization instance\n",
        "optimizer = BayesianOptimization(\n",
        "    f=pipeline,\n",
        "    pbounds=pbounds,\n",
        "    verbose=2,\n",
        "    random_state=1,\n",
        ")\n",
        "# Perform Bayesian optimization\n",
        "optimizer.maximize(init_points=10, n_iter=10)\n",
        "\n",
        "# Print the best set of parameters and the corresponding maximum accuracy score\n",
        "print(optimizer.max)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gqLJDXA4ja_m"
      },
      "source": [
        "# Préparation du dataset pour l'entrainement avec les hyperparamètres optimaux"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8_sm1mSNjmRd"
      },
      "source": [
        "## Feature selection"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 478
        },
        "id": "l7dB2424jsyy",
        "outputId": "00bd7975-e883-40c8-d1cd-e973b9e1706c"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-1f7afc19-2c52-407c-a9b1-0bc534d88af5\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>MILLISEC</th>\n",
              "      <th>Accelerometer RKN^ accX</th>\n",
              "      <th>Accelerometer RKN^ accY</th>\n",
              "      <th>Accelerometer HIP accX</th>\n",
              "      <th>InertialMeasurementUnit BACK gyroX</th>\n",
              "      <th>InertialMeasurementUnit L-SHOE Nav_Ax</th>\n",
              "      <th>InertialMeasurementUnit L-SHOE Nav_Ay</th>\n",
              "      <th>InertialMeasurementUnit L-SHOE Nav_Az</th>\n",
              "      <th>InertialMeasurementUnit L-SHOE AngVelBodyFrameZ</th>\n",
              "      <th>InertialMeasurementUnit R-SHOE Nav_Ax</th>\n",
              "      <th>...</th>\n",
              "      <th>Accelerometer DOOR2 accY</th>\n",
              "      <th>Accelerometer UPPERDRAWER accX</th>\n",
              "      <th>Accelerometer LOWERDRAWER accX</th>\n",
              "      <th>Locomotion</th>\n",
              "      <th>HL_Activity</th>\n",
              "      <th>LL_Left_Arm</th>\n",
              "      <th>LL_Left_Arm_Object</th>\n",
              "      <th>LL_Right_Arm</th>\n",
              "      <th>LL_Right_Arm_Object</th>\n",
              "      <th>ML_Both_Arms</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>-1.585284</td>\n",
              "      <td>0.146747</td>\n",
              "      <td>0.337532</td>\n",
              "      <td>0.693260</td>\n",
              "      <td>-0.221680</td>\n",
              "      <td>0.600405</td>\n",
              "      <td>0.094571</td>\n",
              "      <td>-0.312776</td>\n",
              "      <td>2.280966</td>\n",
              "      <td>0.024974</td>\n",
              "      <td>...</td>\n",
              "      <td>-0.034580</td>\n",
              "      <td>-0.528598</td>\n",
              "      <td>-0.321718</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>-1.585185</td>\n",
              "      <td>0.263802</td>\n",
              "      <td>0.342378</td>\n",
              "      <td>0.626338</td>\n",
              "      <td>-0.228620</td>\n",
              "      <td>-0.196844</td>\n",
              "      <td>-0.040467</td>\n",
              "      <td>0.153620</td>\n",
              "      <td>1.707885</td>\n",
              "      <td>0.017984</td>\n",
              "      <td>...</td>\n",
              "      <td>0.034665</td>\n",
              "      <td>-0.528598</td>\n",
              "      <td>-0.321718</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>-1.585083</td>\n",
              "      <td>0.194202</td>\n",
              "      <td>0.371457</td>\n",
              "      <td>0.470187</td>\n",
              "      <td>-0.226885</td>\n",
              "      <td>-0.229964</td>\n",
              "      <td>-0.031464</td>\n",
              "      <td>-0.014929</td>\n",
              "      <td>1.105162</td>\n",
              "      <td>0.024974</td>\n",
              "      <td>...</td>\n",
              "      <td>1.558045</td>\n",
              "      <td>-0.528598</td>\n",
              "      <td>-0.321718</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>-1.584985</td>\n",
              "      <td>0.058164</td>\n",
              "      <td>0.153369</td>\n",
              "      <td>0.618902</td>\n",
              "      <td>-0.107165</td>\n",
              "      <td>-0.222867</td>\n",
              "      <td>0.065313</td>\n",
              "      <td>-0.040327</td>\n",
              "      <td>0.539986</td>\n",
              "      <td>0.034294</td>\n",
              "      <td>...</td>\n",
              "      <td>0.657866</td>\n",
              "      <td>-0.528598</td>\n",
              "      <td>-0.063922</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>-1.584886</td>\n",
              "      <td>0.247984</td>\n",
              "      <td>0.290684</td>\n",
              "      <td>0.570570</td>\n",
              "      <td>-0.081138</td>\n",
              "      <td>-0.608480</td>\n",
              "      <td>-0.234021</td>\n",
              "      <td>0.040484</td>\n",
              "      <td>0.313718</td>\n",
              "      <td>0.017984</td>\n",
              "      <td>...</td>\n",
              "      <td>0.242398</td>\n",
              "      <td>-0.528598</td>\n",
              "      <td>-0.321718</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606043</th>\n",
              "      <td>1.407215</td>\n",
              "      <td>0.640278</td>\n",
              "      <td>0.298761</td>\n",
              "      <td>0.358651</td>\n",
              "      <td>0.029907</td>\n",
              "      <td>-0.078558</td>\n",
              "      <td>-0.022462</td>\n",
              "      <td>0.209033</td>\n",
              "      <td>-0.058785</td>\n",
              "      <td>0.004005</td>\n",
              "      <td>...</td>\n",
              "      <td>-1.073249</td>\n",
              "      <td>-0.160163</td>\n",
              "      <td>-0.321718</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606044</th>\n",
              "      <td>1.407317</td>\n",
              "      <td>0.437804</td>\n",
              "      <td>0.297145</td>\n",
              "      <td>0.291729</td>\n",
              "      <td>0.002146</td>\n",
              "      <td>-0.111678</td>\n",
              "      <td>-0.074226</td>\n",
              "      <td>0.271373</td>\n",
              "      <td>-0.176365</td>\n",
              "      <td>0.024974</td>\n",
              "      <td>...</td>\n",
              "      <td>0.242398</td>\n",
              "      <td>0.899089</td>\n",
              "      <td>-0.321718</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606045</th>\n",
              "      <td>1.407416</td>\n",
              "      <td>0.678242</td>\n",
              "      <td>0.240604</td>\n",
              "      <td>0.143014</td>\n",
              "      <td>0.092370</td>\n",
              "      <td>-0.222867</td>\n",
              "      <td>-0.047219</td>\n",
              "      <td>0.301389</td>\n",
              "      <td>-0.114117</td>\n",
              "      <td>0.013324</td>\n",
              "      <td>...</td>\n",
              "      <td>-0.311558</td>\n",
              "      <td>0.254327</td>\n",
              "      <td>-0.321718</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606046</th>\n",
              "      <td>1.407514</td>\n",
              "      <td>1.155955</td>\n",
              "      <td>0.159831</td>\n",
              "      <td>0.098399</td>\n",
              "      <td>0.113191</td>\n",
              "      <td>-0.158992</td>\n",
              "      <td>-0.060722</td>\n",
              "      <td>0.174400</td>\n",
              "      <td>-0.125973</td>\n",
              "      <td>0.001675</td>\n",
              "      <td>...</td>\n",
              "      <td>-0.311558</td>\n",
              "      <td>0.162218</td>\n",
              "      <td>0.022009</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606047</th>\n",
              "      <td>1.407616</td>\n",
              "      <td>0.323912</td>\n",
              "      <td>0.445768</td>\n",
              "      <td>0.131860</td>\n",
              "      <td>0.166978</td>\n",
              "      <td>-0.040706</td>\n",
              "      <td>-0.116988</td>\n",
              "      <td>0.088971</td>\n",
              "      <td>-0.228733</td>\n",
              "      <td>0.008665</td>\n",
              "      <td>...</td>\n",
              "      <td>-1.073249</td>\n",
              "      <td>-0.528598</td>\n",
              "      <td>-3.673057</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>606048 rows × 39 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-1f7afc19-2c52-407c-a9b1-0bc534d88af5')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-1f7afc19-2c52-407c-a9b1-0bc534d88af5 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-1f7afc19-2c52-407c-a9b1-0bc534d88af5');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-3b52e49f-135c-49e2-b30d-cd6e0ffeb44f\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-3b52e49f-135c-49e2-b30d-cd6e0ffeb44f')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-3b52e49f-135c-49e2-b30d-cd6e0ffeb44f button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "text/plain": [
              "        MILLISEC  Accelerometer RKN^ accX  Accelerometer RKN^ accY  \\\n",
              "0      -1.585284                 0.146747                 0.337532   \n",
              "1      -1.585185                 0.263802                 0.342378   \n",
              "2      -1.585083                 0.194202                 0.371457   \n",
              "3      -1.584985                 0.058164                 0.153369   \n",
              "4      -1.584886                 0.247984                 0.290684   \n",
              "...          ...                      ...                      ...   \n",
              "606043  1.407215                 0.640278                 0.298761   \n",
              "606044  1.407317                 0.437804                 0.297145   \n",
              "606045  1.407416                 0.678242                 0.240604   \n",
              "606046  1.407514                 1.155955                 0.159831   \n",
              "606047  1.407616                 0.323912                 0.445768   \n",
              "\n",
              "        Accelerometer HIP accX  InertialMeasurementUnit BACK gyroX  \\\n",
              "0                     0.693260                           -0.221680   \n",
              "1                     0.626338                           -0.228620   \n",
              "2                     0.470187                           -0.226885   \n",
              "3                     0.618902                           -0.107165   \n",
              "4                     0.570570                           -0.081138   \n",
              "...                        ...                                 ...   \n",
              "606043                0.358651                            0.029907   \n",
              "606044                0.291729                            0.002146   \n",
              "606045                0.143014                            0.092370   \n",
              "606046                0.098399                            0.113191   \n",
              "606047                0.131860                            0.166978   \n",
              "\n",
              "        InertialMeasurementUnit L-SHOE Nav_Ax  \\\n",
              "0                                    0.600405   \n",
              "1                                   -0.196844   \n",
              "2                                   -0.229964   \n",
              "3                                   -0.222867   \n",
              "4                                   -0.608480   \n",
              "...                                       ...   \n",
              "606043                              -0.078558   \n",
              "606044                              -0.111678   \n",
              "606045                              -0.222867   \n",
              "606046                              -0.158992   \n",
              "606047                              -0.040706   \n",
              "\n",
              "        InertialMeasurementUnit L-SHOE Nav_Ay  \\\n",
              "0                                    0.094571   \n",
              "1                                   -0.040467   \n",
              "2                                   -0.031464   \n",
              "3                                    0.065313   \n",
              "4                                   -0.234021   \n",
              "...                                       ...   \n",
              "606043                              -0.022462   \n",
              "606044                              -0.074226   \n",
              "606045                              -0.047219   \n",
              "606046                              -0.060722   \n",
              "606047                              -0.116988   \n",
              "\n",
              "        InertialMeasurementUnit L-SHOE Nav_Az  \\\n",
              "0                                   -0.312776   \n",
              "1                                    0.153620   \n",
              "2                                   -0.014929   \n",
              "3                                   -0.040327   \n",
              "4                                    0.040484   \n",
              "...                                       ...   \n",
              "606043                               0.209033   \n",
              "606044                               0.271373   \n",
              "606045                               0.301389   \n",
              "606046                               0.174400   \n",
              "606047                               0.088971   \n",
              "\n",
              "        InertialMeasurementUnit L-SHOE AngVelBodyFrameZ  \\\n",
              "0                                              2.280966   \n",
              "1                                              1.707885   \n",
              "2                                              1.105162   \n",
              "3                                              0.539986   \n",
              "4                                              0.313718   \n",
              "...                                                 ...   \n",
              "606043                                        -0.058785   \n",
              "606044                                        -0.176365   \n",
              "606045                                        -0.114117   \n",
              "606046                                        -0.125973   \n",
              "606047                                        -0.228733   \n",
              "\n",
              "        InertialMeasurementUnit R-SHOE Nav_Ax  ...  Accelerometer DOOR2 accY  \\\n",
              "0                                    0.024974  ...                 -0.034580   \n",
              "1                                    0.017984  ...                  0.034665   \n",
              "2                                    0.024974  ...                  1.558045   \n",
              "3                                    0.034294  ...                  0.657866   \n",
              "4                                    0.017984  ...                  0.242398   \n",
              "...                                       ...  ...                       ...   \n",
              "606043                               0.004005  ...                 -1.073249   \n",
              "606044                               0.024974  ...                  0.242398   \n",
              "606045                               0.013324  ...                 -0.311558   \n",
              "606046                               0.001675  ...                 -0.311558   \n",
              "606047                               0.008665  ...                 -1.073249   \n",
              "\n",
              "        Accelerometer UPPERDRAWER accX  Accelerometer LOWERDRAWER accX  \\\n",
              "0                            -0.528598                       -0.321718   \n",
              "1                            -0.528598                       -0.321718   \n",
              "2                            -0.528598                       -0.321718   \n",
              "3                            -0.528598                       -0.063922   \n",
              "4                            -0.528598                       -0.321718   \n",
              "...                                ...                             ...   \n",
              "606043                       -0.160163                       -0.321718   \n",
              "606044                        0.899089                       -0.321718   \n",
              "606045                        0.254327                       -0.321718   \n",
              "606046                        0.162218                        0.022009   \n",
              "606047                       -0.528598                       -3.673057   \n",
              "\n",
              "        Locomotion  HL_Activity  LL_Left_Arm  LL_Left_Arm_Object  \\\n",
              "0                0            0            0                   0   \n",
              "1                0            0            0                   0   \n",
              "2                0            0            0                   0   \n",
              "3                0            0            0                   0   \n",
              "4                0            0            0                   0   \n",
              "...            ...          ...          ...                 ...   \n",
              "606043           0            0            0                   0   \n",
              "606044           0            0            0                   0   \n",
              "606045           0            0            0                   0   \n",
              "606046           0            0            0                   0   \n",
              "606047           0            0            0                   0   \n",
              "\n",
              "        LL_Right_Arm  LL_Right_Arm_Object  ML_Both_Arms  \n",
              "0                  0                    0             0  \n",
              "1                  0                    0             0  \n",
              "2                  0                    0             0  \n",
              "3                  0                    0             0  \n",
              "4                  0                    0             0  \n",
              "...              ...                  ...           ...  \n",
              "606043             0                    0             0  \n",
              "606044             0                    0             0  \n",
              "606045             0                    0             0  \n",
              "606046             0                    0             0  \n",
              "606047             0                    0             0  \n",
              "\n",
              "[606048 rows x 39 columns]"
            ]
          },
          "execution_count": 13,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Get the best set of parameters\n",
        "method = optimizer.max['params']['method']\n",
        "tresh = optimizer.max['params']['tresh']\n",
        "\n",
        "# Apply the best feature selection method for this scenario\n",
        "if method <= 1:\n",
        "  dataset = dim_reduction(df[df.columns[:-7]], tresh)\n",
        "elif method <= 2:\n",
        "  dataset = correlation(df[df.columns[:-7]], tresh)\n",
        "\n",
        "# Concatenate the selected features with the labels\n",
        "labels = df[df.columns[-7:]]\n",
        "dataset = pd.concat([dataset, labels], axis=1)\n",
        "\n",
        "dataset"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5m8XjIrPj4T1"
      },
      "source": [
        "## Segmentation du dataset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FPdPaJFLj8k_",
        "outputId": "74c2cc47-f11f-49e8-b2af-24ce3e044127"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "((4122, 25, 39), (4122,))"
            ]
          },
          "execution_count": 14,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "X, y = data_segmentation(data_collection=dataset, window_size=25, label='HL_Activity')\n",
        "\n",
        "X.shape, y.shape"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eB-ftj3XkEMZ"
      },
      "source": [
        "## Division des données en train, val et test sets"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gMz0CaEzODOa",
        "outputId": "b81acd6c-60f7-4cb1-bb96-651bb91c5eef"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Shape of the train set:  (2761, 25, 39)\n",
            "Shape of the validation set:  (681, 25, 39)\n",
            "Shape of the test set:  (680, 25, 39)\n",
            "Shape of the train set:  (2761, 5)\n",
            "Shape of the validation set:  (681, 5)\n",
            "Shape of the test set:  (680, 5)\n"
          ]
        }
      ],
      "source": [
        "# categorize the data\n",
        "y = to_categorical(y)\n",
        "\n",
        "# split the data into train, validation and test sets using sklearn\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, shuffle=True, test_size=0.33)\n",
        "X_test, X_val, y_test, y_val = train_test_split(X_test, y_test, shuffle=True, test_size=0.5)\n",
        "\n",
        "# display the shape of the train, validation and test sets\n",
        "print(\"Shape of the train set: \", X_train.shape)\n",
        "print(\"Shape of the validation set: \", X_val.shape)\n",
        "print(\"Shape of the test set: \", X_test.shape)\n",
        "\n",
        "# display the shape of the train, validation and test sets\n",
        "print(\"Shape of the train set: \", y_train.shape)\n",
        "print(\"Shape of the validation set: \", y_val.shape)\n",
        "print(\"Shape of the test set: \", y_test.shape)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "slwKyC4ikWAE"
      },
      "source": [
        "# Entrainement du dataset avec les hyperparamètres optimaux"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qoD3_bW5ck_v",
        "outputId": "e4a9a4e2-1705-4265-cdf5-09812e3db159"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "6/6 - 11s - loss: 3.0500 - tp: 1000.0000 - fp: 1092.0000 - tn: 9952.0000 - fn: 1761.0000 - accuracy: 0.4064 - precision: 0.4780 - recall: 0.3622 - auc: 0.6628 - val_loss: 1.3039 - val_tp: 425.0000 - val_fp: 234.0000 - val_tn: 2490.0000 - val_fn: 256.0000 - val_accuracy: 0.6432 - val_precision: 0.6449 - val_recall: 0.6241 - val_auc: 0.8747 - 11s/epoch - 2s/step\n",
            "Epoch 2/100\n",
            "6/6 - 0s - loss: 1.0480 - tp: 1814.0000 - fp: 730.0000 - tn: 10314.0000 - fn: 947.0000 - accuracy: 0.6777 - precision: 0.7131 - recall: 0.6570 - auc: 0.9026 - val_loss: 0.6137 - val_tp: 489.0000 - val_fp: 78.0000 - val_tn: 2646.0000 - val_fn: 192.0000 - val_accuracy: 0.8076 - val_precision: 0.8624 - val_recall: 0.7181 - val_auc: 0.9556 - 332ms/epoch - 55ms/step\n",
            "Epoch 3/100\n",
            "6/6 - 0s - loss: 0.5705 - tp: 1996.0000 - fp: 406.0000 - tn: 10638.0000 - fn: 765.0000 - accuracy: 0.7711 - precision: 0.8310 - recall: 0.7229 - auc: 0.9597 - val_loss: 0.3913 - val_tp: 557.0000 - val_fp: 95.0000 - val_tn: 2629.0000 - val_fn: 124.0000 - val_accuracy: 0.8326 - val_precision: 0.8543 - val_recall: 0.8179 - val_auc: 0.9821 - 229ms/epoch - 38ms/step\n",
            "Epoch 4/100\n",
            "6/6 - 0s - loss: 0.3619 - tp: 2291.0000 - fp: 303.0000 - tn: 10741.0000 - fn: 470.0000 - accuracy: 0.8540 - precision: 0.8832 - recall: 0.8298 - auc: 0.9829 - val_loss: 0.2237 - val_tp: 610.0000 - val_fp: 44.0000 - val_tn: 2680.0000 - val_fn: 71.0000 - val_accuracy: 0.9251 - val_precision: 0.9327 - val_recall: 0.8957 - val_auc: 0.9950 - 225ms/epoch - 37ms/step\n",
            "Epoch 5/100\n",
            "6/6 - 0s - loss: 0.2130 - tp: 2512.0000 - fp: 179.0000 - tn: 10865.0000 - fn: 249.0000 - accuracy: 0.9250 - precision: 0.9335 - recall: 0.9098 - auc: 0.9935 - val_loss: 0.3471 - val_tp: 590.0000 - val_fp: 86.0000 - val_tn: 2638.0000 - val_fn: 91.0000 - val_accuracy: 0.8737 - val_precision: 0.8728 - val_recall: 0.8664 - val_auc: 0.9865 - 219ms/epoch - 36ms/step\n",
            "Epoch 6/100\n",
            "6/6 - 0s - loss: 0.3002 - tp: 2467.0000 - fp: 287.0000 - tn: 10757.0000 - fn: 294.0000 - accuracy: 0.8957 - precision: 0.8958 - recall: 0.8935 - auc: 0.9888 - val_loss: 0.1262 - val_tp: 644.0000 - val_fp: 36.0000 - val_tn: 2688.0000 - val_fn: 37.0000 - val_accuracy: 0.9457 - val_precision: 0.9471 - val_recall: 0.9457 - val_auc: 0.9978 - 223ms/epoch - 37ms/step\n",
            "Epoch 7/100\n",
            "6/6 - 0s - loss: 0.1556 - tp: 2569.0000 - fp: 156.0000 - tn: 10888.0000 - fn: 192.0000 - accuracy: 0.9373 - precision: 0.9428 - recall: 0.9305 - auc: 0.9966 - val_loss: 0.0962 - val_tp: 657.0000 - val_fp: 23.0000 - val_tn: 2701.0000 - val_fn: 24.0000 - val_accuracy: 0.9648 - val_precision: 0.9662 - val_recall: 0.9648 - val_auc: 0.9988 - 256ms/epoch - 43ms/step\n",
            "Epoch 8/100\n",
            "6/6 - 0s - loss: 0.0953 - tp: 2657.0000 - fp: 99.0000 - tn: 10945.0000 - fn: 104.0000 - accuracy: 0.9631 - precision: 0.9641 - recall: 0.9623 - auc: 0.9986 - val_loss: 0.0399 - val_tp: 673.0000 - val_fp: 8.0000 - val_tn: 2716.0000 - val_fn: 8.0000 - val_accuracy: 0.9883 - val_precision: 0.9883 - val_recall: 0.9883 - val_auc: 0.9998 - 239ms/epoch - 40ms/step\n",
            "Epoch 9/100\n",
            "6/6 - 0s - loss: 0.0469 - tp: 2713.0000 - fp: 48.0000 - tn: 10996.0000 - fn: 48.0000 - accuracy: 0.9826 - precision: 0.9826 - recall: 0.9826 - auc: 0.9997 - val_loss: 0.0266 - val_tp: 676.0000 - val_fp: 5.0000 - val_tn: 2719.0000 - val_fn: 5.0000 - val_accuracy: 0.9927 - val_precision: 0.9927 - val_recall: 0.9927 - val_auc: 0.9999 - 227ms/epoch - 38ms/step\n",
            "Epoch 10/100\n",
            "6/6 - 0s - loss: 0.0197 - tp: 2746.0000 - fp: 15.0000 - tn: 11029.0000 - fn: 15.0000 - accuracy: 0.9946 - precision: 0.9946 - recall: 0.9946 - auc: 1.0000 - val_loss: 0.0108 - val_tp: 678.0000 - val_fp: 2.0000 - val_tn: 2722.0000 - val_fn: 3.0000 - val_accuracy: 0.9956 - val_precision: 0.9971 - val_recall: 0.9956 - val_auc: 1.0000 - 227ms/epoch - 38ms/step\n",
            "Epoch 11/100\n",
            "6/6 - 0s - loss: 0.0044 - tp: 2757.0000 - fp: 4.0000 - tn: 11040.0000 - fn: 4.0000 - accuracy: 0.9986 - precision: 0.9986 - recall: 0.9986 - auc: 1.0000 - val_loss: 0.0035 - val_tp: 680.0000 - val_fp: 1.0000 - val_tn: 2723.0000 - val_fn: 1.0000 - val_accuracy: 0.9985 - val_precision: 0.9985 - val_recall: 0.9985 - val_auc: 1.0000 - 262ms/epoch - 44ms/step\n",
            "Epoch 12/100\n",
            "6/6 - 0s - loss: 0.0048 - tp: 2757.0000 - fp: 4.0000 - tn: 11040.0000 - fn: 4.0000 - accuracy: 0.9986 - precision: 0.9986 - recall: 0.9986 - auc: 1.0000 - val_loss: 0.0090 - val_tp: 679.0000 - val_fp: 2.0000 - val_tn: 2722.0000 - val_fn: 2.0000 - val_accuracy: 0.9971 - val_precision: 0.9971 - val_recall: 0.9971 - val_auc: 1.0000 - 264ms/epoch - 44ms/step\n",
            "Epoch 13/100\n",
            "6/6 - 0s - loss: 0.0057 - tp: 2758.0000 - fp: 3.0000 - tn: 11041.0000 - fn: 3.0000 - accuracy: 0.9989 - precision: 0.9989 - recall: 0.9989 - auc: 0.9998 - val_loss: 0.0076 - val_tp: 679.0000 - val_fp: 2.0000 - val_tn: 2722.0000 - val_fn: 2.0000 - val_accuracy: 0.9971 - val_precision: 0.9971 - val_recall: 0.9971 - val_auc: 1.0000 - 256ms/epoch - 43ms/step\n",
            "Epoch 14/100\n",
            "6/6 - 0s - loss: 0.0066 - tp: 2755.0000 - fp: 6.0000 - tn: 11038.0000 - fn: 6.0000 - accuracy: 0.9978 - precision: 0.9978 - recall: 0.9978 - auc: 0.9998 - val_loss: 0.0040 - val_tp: 680.0000 - val_fp: 1.0000 - val_tn: 2723.0000 - val_fn: 1.0000 - val_accuracy: 0.9985 - val_precision: 0.9985 - val_recall: 0.9985 - val_auc: 1.0000 - 264ms/epoch - 44ms/step\n",
            "Epoch 15/100\n",
            "6/6 - 0s - loss: 0.0019 - tp: 2760.0000 - fp: 1.0000 - tn: 11043.0000 - fn: 1.0000 - accuracy: 0.9996 - precision: 0.9996 - recall: 0.9996 - auc: 1.0000 - val_loss: 0.0057 - val_tp: 680.0000 - val_fp: 1.0000 - val_tn: 2723.0000 - val_fn: 1.0000 - val_accuracy: 0.9985 - val_precision: 0.9985 - val_recall: 0.9985 - val_auc: 1.0000 - 259ms/epoch - 43ms/step\n",
            "Epoch 16/100\n",
            "6/6 - 0s - loss: 0.0019 - tp: 2759.0000 - fp: 2.0000 - tn: 11042.0000 - fn: 2.0000 - accuracy: 0.9993 - precision: 0.9993 - recall: 0.9993 - auc: 1.0000 - val_loss: 2.7268e-04 - val_tp: 681.0000 - val_fp: 0.0000e+00 - val_tn: 2724.0000 - val_fn: 0.0000e+00 - val_accuracy: 1.0000 - val_precision: 1.0000 - val_recall: 1.0000 - val_auc: 1.0000 - 263ms/epoch - 44ms/step\n",
            "Epoch 17/100\n",
            "6/6 - 0s - loss: 7.7889e-04 - tp: 2761.0000 - fp: 0.0000e+00 - tn: 11044.0000 - fn: 0.0000e+00 - accuracy: 1.0000 - precision: 1.0000 - recall: 1.0000 - auc: 1.0000 - val_loss: 8.9111e-04 - val_tp: 681.0000 - val_fp: 0.0000e+00 - val_tn: 2724.0000 - val_fn: 0.0000e+00 - val_accuracy: 1.0000 - val_precision: 1.0000 - val_recall: 1.0000 - val_auc: 1.0000 - 230ms/epoch - 38ms/step\n",
            "Epoch 18/100\n",
            "6/6 - 0s - loss: 0.0026 - tp: 2760.0000 - fp: 1.0000 - tn: 11043.0000 - fn: 1.0000 - accuracy: 0.9996 - precision: 0.9996 - recall: 0.9996 - auc: 0.9998 - val_loss: 5.2282e-04 - val_tp: 681.0000 - val_fp: 0.0000e+00 - val_tn: 2724.0000 - val_fn: 0.0000e+00 - val_accuracy: 1.0000 - val_precision: 1.0000 - val_recall: 1.0000 - val_auc: 1.0000 - 256ms/epoch - 43ms/step\n",
            "Epoch 19/100\n",
            "6/6 - 0s - loss: 0.0015 - tp: 2760.0000 - fp: 1.0000 - tn: 11043.0000 - fn: 1.0000 - accuracy: 0.9996 - precision: 0.9996 - recall: 0.9996 - auc: 1.0000 - val_loss: 2.4247e-04 - val_tp: 681.0000 - val_fp: 0.0000e+00 - val_tn: 2724.0000 - val_fn: 0.0000e+00 - val_accuracy: 1.0000 - val_precision: 1.0000 - val_recall: 1.0000 - val_auc: 1.0000 - 289ms/epoch - 48ms/step\n",
            "Epoch 20/100\n",
            "6/6 - 0s - loss: 0.0032 - tp: 2756.0000 - fp: 5.0000 - tn: 11039.0000 - fn: 5.0000 - accuracy: 0.9982 - precision: 0.9982 - recall: 0.9982 - auc: 1.0000 - val_loss: 4.8075e-04 - val_tp: 681.0000 - val_fp: 0.0000e+00 - val_tn: 2724.0000 - val_fn: 0.0000e+00 - val_accuracy: 1.0000 - val_precision: 1.0000 - val_recall: 1.0000 - val_auc: 1.0000 - 276ms/epoch - 46ms/step\n",
            "Epoch 21/100\n",
            "6/6 - 0s - loss: 0.0025 - tp: 2759.0000 - fp: 2.0000 - tn: 11042.0000 - fn: 2.0000 - accuracy: 0.9993 - precision: 0.9993 - recall: 0.9993 - auc: 1.0000 - val_loss: 5.8469e-04 - val_tp: 681.0000 - val_fp: 0.0000e+00 - val_tn: 2724.0000 - val_fn: 0.0000e+00 - val_accuracy: 1.0000 - val_precision: 1.0000 - val_recall: 1.0000 - val_auc: 1.0000 - 261ms/epoch - 43ms/step\n",
            "Epoch 22/100\n",
            "6/6 - 0s - loss: 0.0026 - tp: 2759.0000 - fp: 2.0000 - tn: 11042.0000 - fn: 2.0000 - accuracy: 0.9993 - precision: 0.9993 - recall: 0.9993 - auc: 1.0000 - val_loss: 4.8252e-04 - val_tp: 681.0000 - val_fp: 0.0000e+00 - val_tn: 2724.0000 - val_fn: 0.0000e+00 - val_accuracy: 1.0000 - val_precision: 1.0000 - val_recall: 1.0000 - val_auc: 1.0000 - 294ms/epoch - 49ms/step\n",
            "Epoch 23/100\n",
            "6/6 - 0s - loss: 0.0014 - tp: 2760.0000 - fp: 1.0000 - tn: 11043.0000 - fn: 1.0000 - accuracy: 0.9996 - precision: 0.9996 - recall: 0.9996 - auc: 1.0000 - val_loss: 0.0192 - val_tp: 678.0000 - val_fp: 3.0000 - val_tn: 2721.0000 - val_fn: 3.0000 - val_accuracy: 0.9956 - val_precision: 0.9956 - val_recall: 0.9956 - val_auc: 0.9991 - 260ms/epoch - 43ms/step\n",
            "Epoch 24/100\n",
            "6/6 - 0s - loss: 0.0033 - tp: 2759.0000 - fp: 2.0000 - tn: 11042.0000 - fn: 2.0000 - accuracy: 0.9993 - precision: 0.9993 - recall: 0.9993 - auc: 0.9998 - val_loss: 9.2073e-04 - val_tp: 681.0000 - val_fp: 0.0000e+00 - val_tn: 2724.0000 - val_fn: 0.0000e+00 - val_accuracy: 1.0000 - val_precision: 1.0000 - val_recall: 1.0000 - val_auc: 1.0000 - 294ms/epoch - 49ms/step\n"
          ]
        }
      ],
      "source": [
        "# Get the best set of parameters\n",
        "nbn_l = optimizer.max['params']['nbn_l']\n",
        "nbn_d = optimizer.max['params']['nbn_d']\n",
        "lr = optimizer.max['params']['lr']\n",
        "dp = optimizer.max['params']['dp']\n",
        "bs =optimizer.max['params']['bs']\n",
        "\n",
        "history, model = gru_model(X_train, y_train, X_val, y_val, nbn_l, nbn_d, lr, bs, dp, 2)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QhKe5ywkkcKK"
      },
      "source": [
        "# Evaluation du modèle"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lK9_L7TgQAg3",
        "outputId": "e1691cbd-876f-4d4d-f4ff-7fce75c829a2"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Test loss: 0.01147029921412468\n",
            "Test accuracy: 679.0\n"
          ]
        }
      ],
      "source": [
        "score = model.evaluate(X_test, y_test, verbose=0)\n",
        "print('Test loss:', score[0])\n",
        "print('Test accuracy:', score[1])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "e1jDJRlRQEaB",
        "outputId": "c32cfc99-d1da-4f35-cc09-7d7c53edf067"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.plot(history.history['loss'], label='loss')\n",
        "plt.plot(history.history['val_loss'], label='Validation loss')\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel('loss')\n",
        "plt.legend(['loss', 'val_loss'])\n",
        "plt.title('Training and Validation loss')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "3vZzJMitYDyt",
        "outputId": "7c9ff9fe-bb07-44ec-eef8-374205dd39da"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.plot(history.history['accuracy'], label='Accuracy')\n",
        "plt.plot(history.history['val_accuracy'], label='Validation Accuracy')\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel('accuracy')\n",
        "plt.legend(['accuracy', 'val_accuracy'])\n",
        "plt.title('Training and Validation accuracy')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RK_cjjCkYGJs",
        "outputId": "d691d4a2-be73-47d8-f8f9-e33c4547df21"
      },
      "outputs": [
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAHHCAYAAABDUnkqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABpXElEQVR4nO3dd3xT9f7H8VeStukulJbSQqEsWbKXgFxEUIaiKHpRUYaKiiAiDkRlK/yuA3AguDeKIOJAEUTQCyJ42Qoio1D2pqUtXcn5/REaiC3QpmnTpu/n43EeSU7O+Z5Pkqb99DtNhmEYiIiIiPgIs7cDEBEREfEkJTciIiLiU5TciIiIiE9RciMiIiI+RcmNiIiI+BQlNyIiIuJTlNyIiIiIT1FyIyIiIj5FyY2IiIj4FCU3IpcwcOBAEhIS3Dp3/PjxmEwmzwZUyuzevRuTycT7779f4tc2mUyMHz/e+fj999/HZDKxe/fuS56bkJDAwIEDPRpPUX5WfM1VV13FVVddVahzCvP5iVyMkhsps0wmU4G25cuXezvUcm/48OGYTCZ27NhxwWOefvppTCYTmzZtKsHICu/AgQOMHz+eDRs2eDsUEbkAk9aWkrLq448/dnn84YcfsmTJEj766COX/ddccw0xMTFuXyc7Oxu73Y7Vai30uTk5OeTk5BAYGOj29Uu73bt3U7NmTd57770L1oSsXr2aK664ggkTJjB27Nh8j6lVqxahoaGFSm5MJhPjxo1z1t7YbDays7OxWq2XrDFLSEjgqquuKnSN0//+9z9at26d7+stys+Kr8nKygIgICCgwOcU5vMTuRg/bwcg4q4777zT5fFvv/3GkiVL8uz/p/T0dIKDgwt8HX9/f7fiA/Dz88PPT1+ztm3bUqdOHT799NN8k5tVq1aRmJjI//3f/xXpOhaLBYvFUqQyiqIoPyveVNjvREEUJqnJ5e3PT3yHmqXEp1111VVcfvnlrF27ln/9618EBwfz1FNPAfDVV19x3XXXERcXh9VqpXbt2kyaNAmbzeZSxj/7UeT2MXnxxRd58803qV27NlarldatW/P777+7nJtfnxuTycSwYcNYsGABl19+OVarlUaNGrFo0aI88S9fvpxWrVoRGBhI7dq1eeONNwrcj+e///0vt956K9WrV8dqtRIfH88jjzzCmTNn8ry+0NBQ9u/fT+/evQkNDSU6OprHHnssz3tx6tQpBg4cSEREBBUqVGDAgAGcOnXqkrEA9OvXj7/++ot169bleW727NmYTCZuv/12srKyGDt2LC1btiQiIoKQkBA6duzIsmXLLnmN/PpsGIbBs88+S7Vq1QgODqZz5878+eefec49ceIEjz32GI0bNyY0NJTw8HB69OjBxo0bnccsX76c1q1bAzBo0CBn02du7U9+fW7S0tJ49NFHiY+Px2q1Uq9ePV588UX+WWlemJ+Lf1q+fDkmk4k5c+bw1FNPUaVKFUJCQrjhhhvYu3evy7EX+05kZmYybtw46tSp4/yZeeKJJ8jMzMxzzY8//pg2bdoQHBxMxYoV+de//sXixYtdrvPPPjevvvoqjRo1cp7TqlUrZs+e7Xz+Qn1uXn/9dRo1aoTVaiUuLo6hQ4fm+bnLfV1btmyhc+fOBAcHU7VqVZ5//vlLvn/ie/Qvpfi848eP06NHD2677TbuvPNOZxPV+++/T2hoKCNHjiQ0NJSffvqJsWPHkpKSwgsvvHDJcmfPns3p06e5//77MZlMPP/889x8883s2rXrkv/Br1ixgvnz5/Pggw8SFhbGK6+8Qp8+fUhKSqJSpUoArF+/nu7duxMbG8uECROw2WxMnDiR6OjoAr3uuXPnkp6ezpAhQ6hUqRJr1qzh1VdfZd++fcydO9flWJvNRrdu3Wjbti0vvvgiP/74Iy+99BK1a9dmyJAhgCNJuPHGG1mxYgUPPPAADRo04Msvv2TAgAEFiqdfv35MmDCB2bNn06JFC5drf/7553Ts2JHq1atz7Ngx3n77bW6//XYGDx7M6dOneeedd+jWrRtr1qyhWbNmBbperrFjx/Lss8/Ss2dPevbsybp167j22mudzSa5du3axYIFC7j11lupWbMmhw8f5o033qBTp05s2bKFuLg4GjRowMSJExk7diz33XcfHTt2BKB9+/b5XtswDG644QaWLVvGPffcQ7Nmzfjhhx94/PHH2b9/P9OmTXM5viA/Fxfz3HPPYTKZGDVqFEeOHGH69Ol07dqVDRs2EBQU5Dwuv++E3W7nhhtuYMWKFdx33300aNCAzZs3M23aNP7++28WLFjgPH/ChAmMHz+e9u3bM3HiRAICAli9ejU//fQT1157bb6xvfXWWwwfPpxbbrmFhx9+mIyMDDZt2sTq1au54447Lviaxo8fz4QJE+jatStDhgxh27ZtzJw5k99//52VK1e6fNdOnjxJ9+7dufnmm/n3v//NvHnzGDVqFI0bN6ZHjx6XfP/EhxgiPmLo0KHGP3+kO3XqZADGrFmz8hyfnp6eZ9/9999vBAcHGxkZGc59AwYMMGrUqOF8nJiYaABGpUqVjBMnTjj3f/XVVwZgfPPNN85948aNyxMTYAQEBBg7duxw7tu4caMBGK+++qpzX69evYzg4GBj//79zn3bt283/Pz88pSZn/xe35QpUwyTyWTs2bPH5fUBxsSJE12Obd68udGyZUvn4wULFhiA8fzzzzv35eTkGB07djQA47333rtkTK1btzaqVatm2Gw2575FixYZgPHGG284y8zMzHQ57+TJk0ZMTIxx9913u+wHjHHjxjkfv/feewZgJCYmGoZhGEeOHDECAgKM6667zrDb7c7jnnrqKQMwBgwY4NyXkZHhEpdhOD5rq9Xq8t78/vvvF3y9//xZyX3Pnn32WZfjbrnlFsNkMrn8DBT05yI/y5YtMwCjatWqRkpKinP/559/bgDGyy+/7Nx3oe/ERx99ZJjNZuO///2vy/5Zs2YZgLFy5UrDMBw/g2az2bjpppvyvF/nv8edOnUyOnXq5Hx84403Go0aNbro67jQ53fttde6XOu1114zAOPdd9/N87o+/PBD577MzEyjSpUqRp8+fS56XfE9apYSn2e1Whk0aFCe/ef/J3v69GmOHTtGx44dSU9P56+//rpkuX379qVixYrOx7n/xe/ateuS53bt2pXatWs7Hzdp0oTw8HDnuTabjR9//JHevXsTFxfnPK5OnToF/g/0/NeXlpbGsWPHaN++PYZhsH79+jzHP/DAAy6PO3bs6PJavvvuO/z8/Jw1OeDoI/HQQw8VKB5w9JPat28fv/zyi3Pf7NmzCQgI4NZbb3WWmdtfw263c+LECXJycmjVqlW+TVoX8+OPP5KVlcVDDz3k0pQ3YsSIPMdarVbMZsevRJvNxvHjxwkNDaVevXqFvm6u7777DovFwvDhw132P/rooxiGwffff++y/1I/F5fSv39/wsLCnI9vueUWYmNj+e6771yOy+87MXfuXBo0aED9+vU5duyYc7v66qsBnM2CCxYswG63M3bsWOf7letizaUVKlRg3759eZpuLyb38xsxYoTLtQYPHkx4eDgLFy50OT40NNSlz11AQABt2rQp8PsnvkPJjfi8qlWr5tu58c8//+Smm24iIiKC8PBwoqOjnb8Yk5OTL1lu9erVXR7nJjonT54s9Lm55+eee+TIEc6cOUOdOnXyHJffvvwkJSUxcOBAIiMjnf1oOnXqBOR9fYGBgXmau86PB2DPnj3ExsYSGhrqcly9evUKFA/AbbfdhsVicfazyMjI4Msvv6RHjx4uieIHH3xAkyZNCAwMpFKlSkRHR7Nw4cICfS7n27NnDwB169Z12R8dHe1yPXAkUtOmTaNu3bpYrVaioqKIjo5m06ZNhb7u+dePi4tzSTgAGjRo4BJfrkv9XFzKP1+nyWSiTp06efqw5Ped2L59O3/++SfR0dEu22WXXQY4fiYBdu7cidlspmHDhgWKKdeoUaMIDQ2lTZs21K1bl6FDh7Jy5cqLnpP7/vzzZywgIIBatWrlef+qVauWJ8EqzPsnvkN9bsTnnV+DkevUqVN06tSJ8PBwJk6cSO3atQkMDGTdunWMGjUKu91+yXIvNKrDKMDsCkU5tyBsNhvXXHMNJ06cYNSoUdSvX5+QkBD279/PwIED87y+khqhUrlyZa655hq++OILZsyYwTfffMPp06fp16+f85iPP/6YgQMH0rt3bx5//HEqV66MxWJhypQp7Ny5s9himzx5MmPGjOHuu+9m0qRJREZGYjabGTFiRIF+HjyhuH8ucuX3nbDb7TRu3JipU6fme058fHyRrtmgQQO2bdvGt99+y6JFi/jiiy94/fXXGTt2LBMmTChS2blK6v2T0k/JjZRLy5cv5/jx48yfP59//etfzv2JiYlejOqcypUrExgYmO+kdxebCC/X5s2b+fvvv/nggw/o37+/c/+SJUvcjqlGjRosXbqU1NRUl9qbbdu2Faqcfv36sWjRIr7//ntmz55NeHg4vXr1cj4/b948atWqxfz5813+Cx83bpxbMYOjVqJWrVrO/UePHs3z3/y8efPo3Lkz77zzjsv+U6dOERUV5XxcmPlXatSowY8//sjp06ddam9ymz1z4/OU7du3uzw2DIMdO3bQpEmTS55bu3ZtNm7cSJcuXS76GmvXro3dbmfLli2F7twdEhJC37596du3L1lZWdx8880899xzjB49Ot+5oHLfn23btrl8fllZWSQmJtK1a9dCXV/KDzVLSbmU+x/e+f/RZWVl8frrr3srJBcWi4WuXbuyYMECDhw44Ny/Y8eOPP00LnQ+uL4+wzB4+eWX3Y6pZ8+e5OTkMHPmTOc+m83Gq6++WqhyevfuTXBwMK+//jrff/89N998s8sftvxiX716NatWrSp0zF27dsXf359XX33Vpbzp06fnOdZiseT5D3/u3Lns37/fZV9ISAhAgYbA9+zZE5vNxmuvveayf9q0aZhMJo+P4Pnwww85ffq08/G8efM4ePBgga7z73//m/379/PWW2/lee7MmTOkpaUBjs/PbDYzceLEPDVaF6shOX78uMvjgIAAGjZsiGEYZGdn53tO165dCQgI4JVXXnEp+5133iE5OZnrrrvukq9LyifV3Ei51L59eypWrMiAAQOcSwN89NFHpar6evz48SxevJgOHTowZMgQ5x/Jyy+//JJT/9evX5/atWvz2GOPsX//fsLDw/niiy+K1PegV69edOjQgSeffJLdu3fTsGFD5s+fX+j+KKGhofTu3dvZ7+b8JimA66+/nvnz53PTTTdx3XXXkZiYyKxZs2jYsCGpqamFulbufD1Tpkzh+uuvp2fPnqxfv57vv//epTYm97oTJ05k0KBBtG/fns2bN/PJJ5+41BiAo+aiQoUKzJo1i7CwMEJCQmjbti01a9bMc/1evXrRuXNnnn76aXbv3k3Tpk1ZvHgxX331FSNGjHDpPOwJkZGRXHnllQwaNIjDhw8zffp06tSpw+DBgy957l133cXnn3/OAw88wLJly+jQoQM2m42//vqLzz//nB9++IFWrVpRp04dnn76aSZNmkTHjh25+eabsVqt/P7778TFxTFlypR8y7/22mupUqUKHTp0ICYmhq1bt/Laa69x3XXX5emTlCs6OprRo0czYcIEunfvzg033MC2bdt4/fXXad269SUn7JTyS8mNlEuVKlXi22+/5dFHH+WZZ56hYsWK3HnnnXTp0oVu3bp5OzwAWrZsyffff89jjz3GmDFjiI+PZ+LEiWzduvWSo7n8/f355ptvGD58OFOmTCEwMJCbbrqJYcOG0bRpU7fiMZvNfP3114wYMYKPP/4Yk8nEDTfcwEsvvUTz5s0LVVa/fv2YPXs2sbGxztE4uQYOHMihQ4d44403+OGHH2jYsCEff/wxc+fOdWudsGeffZbAwEBmzZrFsmXLaNu2LYsXL87zX/9TTz1FWloas2fPZs6cObRo0YKFCxfy5JNPuhzn7+/PBx98wOjRo3nggQfIycnhvffeyze5yX3Pxo4dy5w5c3jvvfdISEjghRde4NFHHy30a7mUp556ik2bNjFlyhROnz5Nly5deP311ws0+7DZbGbBggVMmzaNDz/8kC+//JLg4GBq1arFww8/7OxYDDBx4kRq1qzJq6++ytNPP01wcDBNmjThrrvuumD5999/P5988glTp04lNTWVatWqMXz4cJ555pmLxjV+/Hiio6N57bXXeOSRR4iMjOS+++5j8uTJZXZGaCl+WltKpIzp3bs3f/75Z57+FVJ+LV++nM6dOzN37lxuueUWb4cj4nXqcyNSiv1zqYTt27fz3Xff5ZnWXkREzlGzlEgpVqtWLQYOHOic02PmzJkEBATwxBNPeDs0EZFSS8mNSCnWvXt3Pv30Uw4dOoTVaqVdu3ZMnjw5z2RtIiJyjvrciIiIiE9RnxsRERHxKUpuRERExKeUuz43drudAwcOEBYWVqhp1EVERMR7DMPg9OnTxMXF5VmR/p/KXXJz4MCBIi8AJyIiIt6xd+9eqlWrdtFjyl1ykzvN9969ewkPD/dyNCIiIlIQKSkpxMfHX3C5jvOVu+QmtykqPDxcyY2IiEgZU5AuJepQLCIiIj5FyY2IiIj4FCU3IiIi4lOU3IiIiIhPUXIjIiIiPkXJjYiIiPgUJTciIiLiU5TciIiIiE9RciMiIiI+RcmNiIiI+BSvJje//PILvXr1Ii4uDpPJxIIFCy55zvLly2nRogVWq5U6derw/vvvF3ucIiIiUnZ4NblJS0ujadOmzJgxo0DHJyYmct1119G5c2c2bNjAiBEjuPfee/nhhx+KOVIREREpK7y6cGaPHj3o0aNHgY+fNWsWNWvW5KWXXgKgQYMGrFixgmnTptGtW7fiClNEpOyx2yArFbLSwJ7jXhF2g2y7QY7dICfHTrbdTrbNjtlkwmw2YTaZsJhMZx+DxbnPcb8gCxyKm0wWCI0BSylb/9ow4PQhyDkDkbW8FkYpe1cubtWqVXTt2tVlX7du3RgxYsQFz8nMzCQzM9P5OCUlpbjCExFbDmSlkpGewtHjJ8jIMbAHhGD3C8HuHwwm9yuLDQPshoHNbpy95bz7522Ggf3s7fnH/nOfyWQi0M9MoL/l7GYm6Lz7Vr9z9wP9Lfhb/hF7TqYjcchNIAy769M2O2lZNtIzbaRm5ZCemUNaVg6pmbaz922kZWafvc0hLdNGWlYOaZk52LKzCTZlEmLKIJQMgk0ZhJBBMGcIJsOxGRkEcYYgI4Mg4wyBnCHQyCDIfgarcQarkUlRmQHr2U1Knxz8OOZfhcN+VTnsX5WDlqoc8otjvzmOI6YobJjJsZ//fTj3PXD5PtgNDHAkpmbT2ST1XLLqZzmbxJpzk1moaKQQZz9ArG0/cTn7qJJzgJic/URn7yfQyGB3hbYkjFjstfemTCU3hw4dIiYmxmVfTEwMKSkpnDlzhqCgoDznTJkyhQkTJpRUiCJlk90GyXshef/ZP9Zn/2Bnpp73B/zcfSMzlewzp8nJOI2RkYopOw1/Wzr+RhYAgUB8PpdJM6ykE0iaEUgaji3dCCT17G0agaRjJc0Ict5PNYJIJ5B0w4odz9UEBJhyziYMGYScTR5CziYS5ycUIec9H2rKINiUSTAZ+HPx2hA/IOLs5k05hpkcLF6OQjzNgh1/Uw5VsvdRJXsfnHF9PtPwI8mIYbdRhUSjivM20V6Fw1TEKECvlAqcpqbpEAmmQySYD527bzpEuOnMBc+zGSZOp134+ZJQppIbd4wePZqRI0c6H6ekpBAfn9+vXREfZ7dDyn44sROO74QTu87e7oSTu8GWVeCiTEDA2S0/WYaFMwRiNhkEk4EFR61GiCmTEDKJNiUX9dWUGpmGP2lYsV3gj4UJEyYTmE3n7ptMZ/dx7v6558BktmDzC8HuF0yOfwg2v2BsfsFk+4WQYwkmxy+YbEswOZZgsiyO+1mWYLLMQc7bTHMQmeZgsvHHYjHjn7v5mQmwmPC3mAnwO7c/wGLG38907r7FjL/FdPZ4x2OL2TW5NAwDu/GPGrTcWgHnffLZd67WwDA891kYBmTkOGrC0rPOuz1bI5aWaSM9K+dsjdq52rJ/HnuhmPzMJpeaPuftebV8Vn8LgX4WggJc9wf6W/K8f0V7sXbCs44QcSaJiPS9hKcnEZa+h7C0JELSkrCSQ13TfuqyP8+pNksgmeE1yAyvSVZ4AlkVamJYgvBPTiQgORFrSiKBKbvxy7rw99TAxJmgWE6H1OB0SHVSgh1bclB1kq1xREaE0thzr7bQylRyU6VKFQ4fPuyy7/Dhw4SHh+dbawNgtVqxWlWpKuWEYUDKAUfC4kxezt6eTIScjAueajMHcMxSmRR7IKdsVpJtAaQTSKoR6KhtOVvj4tgXRIYpkMDQcMLCK1KxQkWiIiOJjqpEXHQU1aIiiAwJcPS5MAzHdTNda38uWjuU77HpgAf/Epr9wRoKASEQEHp2C3Fs1jAICMHuH0KOX4gzWcgwB5NhCuQMgaSbgkkngAybmYxsO0EBFsIC/QgP9Ccs0I+ws7d5mrMKyN9zr7TYmM7rX+MrDMPgTLaNtEwbWTY71tymSz8zfm5+lsWnNtAu7+7cmth//hNzfCec2oPFlkHwyW0En9x26UuExUGl2o7+M5VqQ2RtqFQHU8UEgv0DCQZiLllIyStTyU27du347rvvXPYtWbKEdu3y+XBFfJVhQOph119YJ3bC8V2OX2Q5F6kONvtDxQSoVBsjshZJplh+OBDCnF0BJGZUwP6P2odKIQFUiwymemQw8RWDqBUZTPzZx7ERgQX7ZW8ygX+QYyO6SC+9pJk5V0MV6uVYpGSYTCaCA/wIDihTfx5dmS2O73nFBKCL63O2bDiVlDfpycmAyJpnk5ezSUxkLQgI9sILKDqvfnqpqans2LHD+TgxMZENGzYQGRlJ9erVGT16NPv37+fDDz8E4IEHHuC1117jiSee4O677+ann37i888/Z+HChd56CSLFwzAg/Xg+CcwOOJHoqMm4EJMFKtZw/SVVqZbjNiKeExl25q/bx5zf97L9yLlyakeHcHOLatStHEr1SsFUqxhMqLUM/4IXkbws/o7fC5VqQ91rvB1NsfHqb67//e9/dO7c2fk4t2/MgAEDeP/99zl48CBJSUnO52vWrMnChQt55JFHePnll6lWrRpvv/22hoFL2ZV+Iu9/ULm1MJkX6ZdiMkNE/HnJy3m3Fao7foGdx243WLXrOJ8u2sTiPw+TZXP0gQn0N3Nd4zhubxNPyxoVNXRXRHyCyTA82Z2r9EtJSSEiIoLk5GTCw8O9HY6UJ4YBm+fBjiXnkpgzJy9yggkiqv2jrfvsbcUa4HfpvmSHUzKYt9ZRS5N0It25//Kq4fRtXZ0bm8URHlgWeneISHlXmL/fqnMWKQnpJ+Cb4bD1m7zP5dthr7ajvdw//47yF5Njs7N821E++30vy7YdwWZ3/P8SZvXjxuZx3Na6OpdX9fYAZRGR4qPkRqS47V4J8wc7hmGb/aHdg1C15dkOezUdo3M8YO+JdOb8vpe5a/dyOOXcBG6tEyrSt3V1rmscS1CA5jsREd+n5EakuNhy4Jfn4ZcXHLPXRtaGW96BuOYeu0Rmjo0lWw7z2Zq9rNhxzLk/MiSAPi2q0rd1PHUqh3nseiIiZYGSG5HicCoJvhgMe39zPG7WD3o875hXxQNsdoPPfk9i6uK/OZ7mmHzPZIIr60RxW+vqXNMwhgC/0jYnh4hIyVByI+Jpf34JXz/sGO1kDYfrp0HjWzxW/OZ9yTzz1R9s3HsKgJhwK/9uFc+/W8UTH1k256QQEfEkJTcinpKVBotGw7oPHI+rtYY+b5+dSKvoks9k89LibXz82x7shqOD8MhrL+OuK2qUwplTRUS8R8mNiCcc2gzz7oZjfwMm6DgSrhqdZ74ZdxiGwYIN+3lu4VaOpTqaoG5sFsfTPRtQOTywyOWLiPgaJTciRWEYsPoNWDLGsfBkWCzc9AbU6uSR4rcfPs0zC/5gdeIJAGpFh/DsjZfTvk6UR8oXEfFFSm5E3JV2DBY8CNt/cDy+rAfcOANCKhW96MwcXvlpO+/8N5Ecu0Ggv5mHrq7L4I611FFYROQSlNyIuGPXcph/P6QeAosVrn0W2gx2DFkqAsMw+OHPQ0z8ZgsHkh0reF/TMIax1zdUZ2ERkQJSciNSGLZs+OlZWPkyYEB0fejzDlS5vMhF7zmexriv/2T5tqMAVKsYxPhejejaMKbIZYuIlCdKbkQK6sQu+OJe2L/W8bjlIOg2GQKKVqOSkW3jjZ93MWP5DrJy7PhbTNz/r9oM7VxHMwqLiLhByY1IQWz6HL4dCVmnITACbngNGt5Q5GJ//vso4776g93HHYtadqhTiYk3Xk7taM9M9iciUh4puRG5mIwU+P4J2Pip43H19nDzm1AhvkjFHkw+w6Rvt/Dd5kMAVA6zMub6hlzfJBZTEfvtiIiUd0puRC5k67fw3eNw+gCYzNBpFHR8DCzuf22ybXbeX7mbaT/+TXqWDYvZxIB2CTxyTV3CAos+J46IiCi5EckreR989wRsW+h4XLEm9H4darQvUrF/HUrhkTkb2XowBYAW1SvwbO/GNIwLL2rEIiJyHiU3IrnsNljzpmM0VFYqmP2gw8Pwr8fBP8j9Yu0G765M5PlF28iy2akY7M/oHg24pWU1zGY1QYmIeJqSGxGAgxvhm4fhwHrH4/i2cP10iGlYtGKTz/Do5xv5dedxAK6uX5n/9GlCdJi1iAGLiMiFKLmR8i0zFZZPgd9eB8MO1gi4Zjy0GAjmos0E/PXGAzzz5WZSMnII8rfwzPUNuKNNdXUYFhEpZkpupPz6+wdY+Cgk73U8bnQTdP8/CKtSpGKTz2Qz7qs/WLDhAABNq0UwrW8zaml4t4hIiVByI+XP6UPw/SjYssDxOKI6XPcSXHZtkYtetfM4j36+gQPJGZhNMOzqujx0dR38LVoPSkSkpCi5kfLDboe178KPEyAzBUwWaDcUrnoSAkKKVHRmjo2XFv/NW//dhWFAjUrBTOvbjBbVK3ooeBERKSglN1I+HP7T0WF43++Ox3EtoNfLENukyEVvO3Sahz9bz1+HTgNwW+t4xlzfkBCrvl4iIt6g377i27LS4Zfn4ddXwZ4DAWHQZSy0vgfMRVu3yTnE+4dtZOXYiQwJ4P9ubsy1jYrWZ0dERIpGyY34rh1LYeFIOLnb8bj+9dDjeYioWuSiDyaf4bG5G1m5Q0O8RURKGyU34ntSj8APT8HmuY7H4VWh5wtQ/zqPFP/NxgM8fXaId6C/mWeua0i/thriLSJSWii5Ed+SegRmdoC0I471oNrcD1c/DdawIhedkpHNuK/+5Mv1+wEN8RYRKa2U3IhvWfu+I7GpWBNueReqtvBIsb/tOs6jn29k/6kzGuItIlLKKbkR32HLcSQ3AJ2f8khik5Vj56Ul23jzl3NDvKf+uxkta2iIt4hIaaXkRnzH9h8gZT8EV4KGN3qkyDEL/mDO/xwzGGuIt4hI2aDf0uI7fn/Hcdv8TvAr+qilFduPMed/ezGZYMYdLejZOLbIZYqISPFThwHxDSd2wc6ljvstBxW5uPSsHJ6cvwmAAe0SlNiIiJQhSm7EN/zvPcdt7S4QWbPIxb34w9/sO3mGqhWCeLxbvSKXJyIiJUfJjZR92Rmw/mPH/db3FLm4dUknee/XRAAm39xYfWxERMoYJTfidcnp2fx5INn9ArZ8BWdOOCbrq9utSLFk5tgYNW8ThgF9WlSj02XRRSpPRERKnpIbKbzkffDxLfD3D0UuKsdm5/a3fuO6V1Yw5/ck9wr539mOxC0HgqVotSwzlu1k+5FUokIDGHN9gyKVJSIi3qHkRgrvv1NhxxL4aphjYcoieP/X3Ww5mALAmAV/snbPycIVcOgP2LsazH7Qon+RYtl6MIXXl+0AYOKNl1MhOKBI5YmIiHcouZHCyUo/t2ZT2pFztSZuOJySwfQftwOQUCmYLJudIR+v5XBKRsELyb1+/esgzP3VuHNsdkZ9sYkcu0G3RjH0uFwre4uIlFVKbqRwtn4NmSmOmhKAFdMhM9Wtop5duJXUzByaV6/ANw9dSb2YMI6czuT+j9aSkW27dAGZp2HT5477rYrWkfjdlYls2pdMWKAfk268XItgioiUYUpupHDWfei47fgYRNaC9GPw+1uFLmbljmN8s/EAZhNMuvFywgL9ebN/SyKC/Nmw9xRjFvyBYRgXL2TTHMhKhUp1oea/3HgxDruPpfHS4r8BGHNdQyqHB7pdloiIeJ+SGym44zthz0rHatst+kOnJx37V74MGSkFLiYrx87Yr/4A4K4ranB51QgAalQKYcYdLTCbYO7afXzw6+4LF2IY8Pu7jvut7gY3a1rsdoNRX2wiM8dOhzqVuLVVNbfKERGR0kPJjRTc+o8ct3W6QkRVaHyLo9bkzElY80aBi3l7xS52Hk0jKtTKyGtdJ8i7sm4UT/V0jFKatHArv+48ln8he1fDkT/BLwia3e7WywH47Pe9rE48QZC/hSk3NVFzlIiID1ByIwVjy4ENsx33m9/luDVb4KqztTe/vgpnTl2ymP2nzvDqUseIpKd61iciyD/PMfdcWZObmlfFZjcY+sk69p7IZ0TW/87W2lzeB4LcW6H7YPIZpny3FYDHutWjeqVgt8oREZHSRcmNFMz2xZB6GEKi4bLu5/Y3ugmiG0BGMvw285LFTPzmT85k22hTM5KbmlfN9xiTycSUmxvTuGoEJ9Ozue+jtaRn5Zw7IO04/Pml437ru916OYZh8MyXf3A6M4dm8RUY2D7BrXJERKT0UXIjBZPbkbjpbeB33vwv59fe/PY6pJ+4YBHLth3hhz8PYzGbLjkiKdDfwht3tSQqNICtB1N4Yt6mcx2MN3wMtiyIbQZVW7r1cr7ZdJClfx3B32Li+VuaYDGrOUpExFcouZFLSznoqLkBaJ7PRHkNboCYyx1DxFfNyLeIjGwb47/+E4C7OyRQr0rYJS8bVyGI1/u1xM9s4ttNB5n18y6w288tkunmOlIn0rKcsQzrXJfLYi4di4iIlB1KbuTSNs4GwwbxV0D0ZXmfN5vhqtGO+6tnOZqN/mHWzzvZczydmHArD3fNp4wLaFMzkvE3NALg+R/+YuMvX8LJRLBGOPrbuGHiN39yIi2LejFhDLmqtltliIhI6aXkRi7OMM6tuN3irgsfV/86qNLEMe/Mr6+4PLXneBqvL98JwJjrGxJayFW277yiBre3qY5hwPHlrzt2NrsdAkIKVQ7AT38dZsEGx/w6/7mlCQF++gqIiPga/WaXi9uzEk7sgoAwaNj7wseZTND5acf9NW9C6lHA0XF3/Nd/kpVj58o6UVzXONatMCbc0Ihu1XLoZKwFIK1x4deROp2RzdNfOubXuefKmjSLr+BWLCIiUropuZGLy+1IfPnNYA29+LGXdYO4FpCdDiunA7B4y2GWbTuKv8XEhBsbuT2PTICfmal1NmAxGayyNeThpenY7ZeYwfgf/rPoLw4mZ1A9MpiR19S79AkiIlImKbmRCztzCrZ85bjfYsCljz+/9ub3t0k/sY+J32wB4L5/1aJ29CWSo4uxZRPyh2Oenc+4hh+3HmH6j38X+PTVu47z8W9JAPxfn8YEBVjcj0VEREo1JTdyYZvnQk4GVG4IVVsU7Jw6XaBaG8jJYMvnE9l/6gxVKwQxrHPdosXy10JIPQQhlel0w0AAXvlpB4v+OHjJUzOybTw5fzMAt7eJp33tqKLFIiIipZqSG7mw3OUWmt9V8LWbTCbo/BQAjQ/OJ4YTjOvVsOg1Jf97x3Hboj83t67F3R1qAjDy841sO3T6oqdO/3E7icfSiAm38mSPBkWLQ0RESj2vJzczZswgISGBwMBA2rZty5o1ay54bHZ2NhMnTqR27doEBgbStGlTFi1aVILRliMHNzo2SwA06VuoU42anfgr4HKspmwmRy/hmoYxRYvl2HZI/MWxYGfLgYBj6YYOdSqRnmVj8If/41R6Vr6nbt6XzFv/3QXAs70b57vcg4iI+BavJjdz5sxh5MiRjBs3jnXr1tG0aVO6devGkSNH8j3+mWee4Y033uDVV19ly5YtPPDAA9x0002sX7++hCMvB9adrbWpfx2EVCrUqd9uPsT41N4AXJ32PabkfUWLJXcdqbrdoEI8AH4WM6/d3oL4yCCSTqTz0KfrybHZXU7Lttl54otN2OwG1zeJLXqSJSIiZYJXk5upU6cyePBgBg0aRMOGDZk1axbBwcG8++67+R7/0Ucf8dRTT9GzZ09q1arFkCFD6NmzJy+99FIJR+7jss/A5s8d91sUbsh1amYOzy7cwm/2huyNaIXJngX/fdH9WLLSYcMnjvv/mJG4YkgAb97ViiB/C//dfoz/LPrL5fk3f9nF1oMpVAj2d04EKCIivs9ryU1WVhZr166la9eu54Ixm+natSurVq3K95zMzEwCAwNd9gUFBbFixYoLXiczM5OUlBSXTS5h6zeOhTAjqkPNqwp16vQlf3M4JZMalYKpfMMEx871H8PJ3e7F8ud8RywVqkPtLnmebhAbzou3NgXgrf8msmD9fgB2HEnl5R+3AzCuV0OiQq3uXV9ERMocryU3x44dw2azERPj2lQQExPDoUOH8j2nW7duTJ06le3bt2O321myZAnz58/n4MELj5iZMmUKERERzi0+Pt6jr8Mn5c5t0/xOx9IKBfTXoRTe+3U34Jh0z1r7SqjVGew58MsL7sXy+9mOxC0HXTCW65rEMrSzYxmFUV9sYuPeU4z6YhNZNjtX1Yumd7P8Vx8XERHf5PUOxYXx8ssvU7duXerXr09AQADDhg1j0KBBmC/yB3j06NEkJyc7t71795ZgxGXQiV2w+7+ACZrdUeDTDMNgzII/sNkNujeqwlX1KjueyJ33ZsOncHxn4WI5sB4OrAOzv2PE1kU8ek09rq5fmcwcO/9+YxVr95wkJMDCczc1dnviQBERKZu8ltxERUVhsVg4fPiwy/7Dhw9TpUqVfM+Jjo5mwYIFpKWlsWfPHv766y9CQ0OpVavWBa9jtVoJDw932eQicteRqn21s/NuQXyxbj+/7z5JkL+Fsb0annsivjXUvdax8GZha29ya20a3gih0Rc91Gw2Mf22ZtSKDiEzx9Gx+Mke9alaIahw1xQRkTLPa8lNQEAALVu2ZOnSpc59drudpUuX0q5du4ueGxgYSNWqVcnJyeGLL77gxhtvLO5wywdbDqw/23m3EB2Jk9OzmfLdVgCGd6lL3D8TitwVwzfNcQzrLogzp2DzPMf9f3QkvpDwQH/e6t+KmHArV9evTL+2NQp2LRER8SmFW57Zw0aOHMmAAQNo1aoVbdq0Yfr06aSlpTFo0CAA+vfvT9WqVZkyZQoAq1evZv/+/TRr1oz9+/czfvx47HY7TzzxhDdfhu/Y8aNjFuDgSlCvZ4FPe3HxNo6nZVGncij3XFkz7wFVWzjK2/Yd/Pwf6PP2pQvd+BnknIHoBlD94snu+WpHh/Lb6C5qihIRKce8mtz07duXo0ePMnbsWA4dOkSzZs1YtGiRs5NxUlKSS3+ajIwMnnnmGXbt2kVoaCg9e/bko48+okKFCl56BT4md0bipreDX0CBTtm8L5mPV+8BYOKNjQjwu0Bl4FVPOpKbzfOg42NQuf6FCzWMc3PbtL6n4LMjn6XERkSkfDMZhlG4pZXLuJSUFCIiIkhOTlb/m/OdPgxTGzj6xjy4+uLJx1l2u8FNM39l495T3NA0jldub37xE+bc6Rhm3ugmuPX9Cx+X+F/44HrwD4FH/4JAfU4iIuVdYf5+l6nRUlKMNn7qSGyqtS5QYgPw2e972bj3FKFWP565rgBrNuX2vfnzSzj854WPy11HqsmtSmxERKTQlNyIoxkot0mqgB2JT6Rl8fwPjhmBR15zGZXDAy9xBhDTyFFrA7B8Sv7HnD7sqN0BaFWwjsQiIiLnU3IjkLQKju9wNAPlJh+X8J/v/+JUejYNYsPp364Qo5I6PQmYHAnMwY15n1//oWPSv2qtIbZJwcsVERE5S8mNnJuR+PKbwRp26cOTTjLnf47JEJ/t3Qg/SyF+jCrXh8a3OO4v+0ftjd0Gaz9w3FetjYiIuEnJTXmXkQx/LnDcL0CTlN1uMP5rR3+ZW1pWo2WNyMJfs9MoMJnh7+9h/9pz+7cvhuS9EFSxwDVIIiIi/6Tkprz744uz88nUdzQFXcK8tfvYtC+ZMKsfo7oXrONxHlF1oUlfx/3za29yZyRu1g/8C9CHR0REJB9Kbso75yKZd11yPpmUjGxnJ+LhXeoSHVaElbb/9TiYLLBjCexd41g1fMePjuda3e1+uSIiUu4puSnPDm12LE5p9oemt13y8FeXbudYaha1okMY0D6haNeuVBua3e64v2wy/O89wHCsIl6pdtHKFhGRck3JTXm27uzw7/o9ISTqoofuOJLKeyt3AzD2+oYXnom4MP71OJj9YNcyWPOmY18B15ESERG5ECU35VV2hmMhS4DmF+9IbBgGk77dQo7doEv9ylxVr7JnYqiYAM3vPBtPOoTFwWU9PFO2iIiUW0puyqu/voWMUxBeDWp3vuihP/11hJ//PkqAxcyY6xt6No6OjzmaxQBaDgCLV5c7ExERH6DkprxydiTuB2bLBQ/LzLEx8dstANx9ZU0SokI8G0eFeOg2Gep0hTb3ebZsEREpl/Rvcnl0cjck/gyYHMOuL+LdFbvZczydymFWhl1dp3jiaXufYxMREfEA1dyUR+s/dtzWugoqXnjphMMpGbz203YAnuxRn1CrcmERESn9lNyUN3YbrP/Ecb/FXRc99D/f/0Valo3m1SvQu1nVEghORESk6JTclDc7lsLpA44lDupff8HD1iWdZP76/QCM79UIs/niE/yJiIiUFmpnKK2S98GGTyH6Mqj5L0cy4gnrz3YkbnIb+OU/w/D560f9u1U1msZX8My1RURESoCSm9Lq5/+cG9FkMkNsM8eQ7VpXQXzbCyYmF5V6BLZ977h/kSap89ePerybm+tHiYiIeImSm9LqiGMNJ0KiIe0oHFjn2P77EvgFQY3255Kdyo3AXIAWxo2fgT0HqraEmEb5HuLR9aNERES8QMlNaXUy0XHbbx6EVoZdy89tqYdh51LHBo4EqGYnR6JTuzNEVMtbnmHA+rPLLTS/cK2NR9ePEhER8QIlN6VRRoqjtgYgsiYERkCzOxybYcCRrWcTnWWwe6Xj2D/mOTaASnUcC1DWugpqdnScv3c1HPsb/IPh8j75XrZY1o8SEREpYUpuSqPcWpvgKEdicj6TCWIaOrZ2D0JOFuz7/Vyys38tHN/h2H5/y9Ffp2pLsGU7zm90EwSG57lksa0fJSIiUsKU3JRGJ3Y5biNrXfpYvwBI6ODYrn4aMpJh9wrYucyR8Bzf7kh+crXIf5HM3PWj/C0mnvH0+lEiIiIlSMlNaVSY5OafAiOg/nWODeDUXsdSC4m/OPrixLfNc0pmjo1JZ9ePuufKWtT09PpRIiIiJUjJTWnkTG5qFr2sCvHQ/E7HdgHvrtjN7uJeP0pERKSEqMdoaXRit+PWnZqbQjpy3vpRo7pr/SgRESn7lNyURkVpliqk/1t0bv2om5pr/SgRESn7lNyUNlnpjrWfoNiTm3VJJ5m/TutHiYiIb1FyU9qc3O24DYzw3HpS+bDbDSacXT/q1pZaP0pERHyHkpvS5vwmKVPx1aTMW7ePjWfXj3qiu9aPEhER36HkprTJTW4qemCk1AWkZGTz/CKtHyUiIr5JyU1pkzs7cTH2t9H6USIi4suU3JQ2xTxSSutHiYiIr9NfttKmGJMbrR8lIiLlgZKb0iQnE5L3Oe4XQ3Kj9aNERKQ8UHJTmpxKAsMO/iEQ6tlaFa0fJSIi5YWSm9KkGIeBv7dS60eJiEj5oOSmNHEmNwkeLTYzx8aMn3YAWj9KRER8n5Kb0uRE8QwDX590itOZOUSHWbV+lIiI+DwlN6VJMY2U+nXncQDa166k9aNERMTnKbkpTYorudlxDHAkNyIiIr5OyU1pYcuBU3sc9z2Y3KRl5rBh7ykA2teO8li5IiIipZWSm9IieS/Yc8BihbA4jxX7++4T5NgNqlUMIj4y2GPlioiIlFZKbkoL54KZCWD23Mey6rz+NiIiIuWBkpvSopgWzDzXmVhNUiIiUj4ouSktimEYeHJ6Nn8cSAagnWpuRESknFByU1o4R0rV9FiRvyUexzCgdnQIMeGBHitXRESkNFNyU1oUwzDwVWqSEhGRckjJTWlgtxdLs9SvOx3z23SooyYpEREpP5TclAanD4AtE8x+EBHvkSKPnM7g78OpmEzQtqaSGxERKT+U3JQGubU2FaqDxTOLWuY2STWMDadiSIBHyhQRESkLlNyUBsXa30a1NiIiUr4ouSkNiiG50fw2IiJSXnk9uZkxYwYJCQkEBgbStm1b1qxZc9Hjp0+fTr169QgKCiI+Pp5HHnmEjIyMEoq2mHg4udl7Ip2kE+lYzCZa14z0SJkiIiJlhVeTmzlz5jBy5EjGjRvHunXraNq0Kd26dePIkSP5Hj979myefPJJxo0bx9atW3nnnXeYM2cOTz31VAlH7mEeHim1apej1qZptQhCrZ7pwyMiIlJWeDW5mTp1KoMHD2bQoEE0bNiQWbNmERwczLvvvpvv8b/++isdOnTgjjvuICEhgWuvvZbbb7/9krU9pZpheLzmRvPbiIhIeea15CYrK4u1a9fStWvXc8GYzXTt2pVVq1ble0779u1Zu3atM5nZtWsX3333HT179iyRmItF2lHITgNMjtFSRWQYhnN+G3UmFhGR8shrbRbHjh3DZrMRExPjsj8mJoa//vor33PuuOMOjh07xpVXXolhGOTk5PDAAw9ctFkqMzOTzMxM5+OUlBTPvABPya21iYgHP2uRi9t5NI3DKZkE+JlpUaNikcsTEREpa7zeobgwli9fzuTJk3n99ddZt24d8+fPZ+HChUyaNOmC50yZMoWIiAjnFh/vmUnyPMbDa0qtOltr06pGRQL9LR4pU0REpCzxWs1NVFQUFouFw4cPu+w/fPgwVapUyfecMWPGcNddd3HvvfcC0LhxY9LS0rjvvvt4+umnMZvz5mqjR49m5MiRzscpKSmlK8HxcH+bXzW/jYiIlHNeq7kJCAigZcuWLF261LnPbrezdOlS2rVrl+856enpeRIYi8VRO2EYRr7nWK1WwsPDXbZSxYPJjd1uOEdKtVNnYhERKae8Ok545MiRDBgwgFatWtGmTRumT59OWloagwYNAqB///5UrVqVKVOmANCrVy+mTp1K8+bNadu2LTt27GDMmDH06tXLmeSUOR5MbrYeSuFUejYhARaaVIsocnkiIiJlkVeTm759+3L06FHGjh3LoUOHaNasGYsWLXJ2Mk5KSnKpqXnmmWcwmUw888wz7N+/n+joaHr16sVzzz3nrZdQdB7sc5M7BLxNzUj8LWWqO5WIiIjHmIwLtef4qJSUFCIiIkhOTvZ+E1X6CXj+bFLz1AEICClScXe//zs//XWEp3s2YPC/PLeUg4iIiLcV5u+3/r33ptyZicNii5zYZNvsrHb2t1FnYhERKb+U3HiTB/vbbNqXTFqWjQrB/jSMLWWdpkVEREqQkhtv8mh/G8f8Nu1qVcJsNhW5PBERkbJKyY03ebDmRvPbiIiIOCi58SYPJTcZ2Tb+t+ckoPltRERE3B4Kvn37dpYtW8aRI0ew2+0uz40dO7bIgZULJ892KK5YtGapdUknycqxUznMSu3oonVMFhERKevcSm7eeusthgwZQlRUFFWqVMFkOtfHw2QyKbkpiIwUx4rgUOQ+N6vOa5I6/7MQEREpj9xKbp599lmee+45Ro0a5el4yo/cWpvgKAgs2mzC5/rbqElKRETErT43J0+e5NZbb/V0LOWLh/rbpGbmsHHvKUDz24iIiICbyc2tt97K4sWLPR1L+eKh5Ob3xBPk2A3iI4OIjwz2QGAiIiJlm1vNUnXq1GHMmDH89ttvNG7cGH9/f5fnhw8f7pHgfJqHkptfz85v076WmqRERETAzeTmzTffJDQ0lJ9//pmff/7Z5TmTyaTkpiBO7HbcFrEzsbO/TR01SYmIiICbyU1iYqKn4yh/PFBzczItiy0HUwD1txEREclV5En8DMOgnC0sXnRZ6XD6gON+EZKb1YnHMQyoWzmUymGBHgpORESkbHM7ufnwww9p3LgxQUFBBAUF0aRJEz766CNPxua7Tu523AZGQFBFt4vRkgsiIiJ5udUsNXXqVMaMGcOwYcPo0KEDACtWrOCBBx7g2LFjPPLIIx4N0uec3yRVhEn3cpMbLbkgIiJyjlvJzauvvsrMmTPp37+/c98NN9xAo0aNGD9+vJKbS/FAf5sjKRnsOJKKyQRX1Ir0UGAiIiJln1vNUgcPHqR9+/Z59rdv356DBw8WOSiflzs7cRGSm1W7HLU2jeLCqRAc4ImoREREfIJbyU2dOnX4/PPP8+yfM2cOdevWLXJQPi+35qYIC2au3HF2fhs1SYmIiLhwq1lqwoQJ9O3bl19++cXZ52blypUsXbo036RH/sEDzVLn+tuoM7GIiMj53Kq56dOnD6tXryYqKooFCxawYMECoqKiWLNmDTfddJOnY/QtOZmQvM9x383kZu+JdPadPIOf2USbBPW3EREROZ9bNTcALVu25OOPP/ZkLOXDqSQw7OAfAqGV3Soid8mFZvEVCLG6/RGKiIj4pAL/ZUxJSSE8PNx5/2Jyj5N8eGAYuOa3ERERubACJzcVK1bk4MGDVK5cmQoVKmDK5w+zYRiYTCZsNptHg/QpJ3JHSrnXmdgwDM1vIyIichEFTm5++uknIiMd/TuWLVtWbAH5PGfNjXvJzc6jqRw9nYnVz0zz6hU8F5eIiIiPKHBy06lTp3zvSyEVcaRUbq1Nq4SKBPpbPBWViIiIz3BrtNSiRYtYsWKF8/GMGTNo1qwZd9xxBydPnvRYcD6piMmN5rcRERG5OLeSm8cff9zZqXjz5s2MHDmSnj17kpiYyMiRIz0aoE+x5cCpPY77biQ3NrvBb7tOAJrfRkRE5ELcGkecmJhIw4YNAfjiiy/o1asXkydPZt26dfTs2dOjAfqU5L1gzwGLFcLiCn361oMpJJ/JJtTqR5OqEcUQoIiISNnnVs1NQEAA6enpAPz4449ce+21AERGRl5ymHi5dn5nYnPh3/rc+W3a1ozEz+LWRyciIuLz3Kq5ufLKKxk5ciQdOnRgzZo1zJkzB4C///6batWqeTRAn1LEBTO15IKIiMilufXv/2uvvYafnx/z5s1j5syZVK1aFYDvv/+e7t27ezRAn5I7x40bC2Zm2+ysSXT0t1FnYhERkQtzq+amevXqfPvtt3n2T5s2rcgB+bQizHGzad8p0rNsVAz2p36VMA8HJiIi4ju0/EJJKsIw8F93nGuSMpvdW7ZBRESkPNDyCyXFbj9v6YXCJzcrz3Ym1pILIiIiF6flF0rK6QNgywSzH0TEF+rUjGwb6/acArRYpoiIyKVo+YWSkltrU6EGWArX1WntnpNk2ezEhFupFRVSDMGJiIj4DrdGS7333nvMnTs3z/65c+fywQcfFDkon1SEzsS589u0rx2Vb3OgiIiInONWcjNlyhSiovL2/ahcuTKTJ08uclA+qSidic/Ob6MmKRERkUtzK7lJSkqiZs28NRA1atQgKSmpyEH5JDeTm9MZ2Wzalwxo8j4REZGCcCu5qVy5Mps2bcqzf+PGjVSqpD/A+XJzpNTvu09gsxvUqBRMtYrBxRCYiIiIb3Erubn99tsZPnw4y5Ytw2azYbPZ+Omnn3j44Ye57bbbPB1j2WcYbtfc5M5voyYpERGRgnFrhuJJkyaxe/duunTpgp+fowi73U7//v3V5yY/aUchOw1MZqhQvVCnnltPSvPbiIiIFIRbyU1AQABz5sxh0qRJbNy4kaCgIBo3bkyNGjU8HZ9vyK21Ca8GftaCn5aWxZaDjtmg29VSzY2IiEhBuJXc5EpISMAwDGrXru2swZF8uDkM/Lddjlqby2JCiQ4reFIkIiJSnrnV5yY9PZ177rmH4OBgGjVq5Bwh9dBDD/F///d/Hg3QJ7jb3+a8+W1ERESkYNxKbkaPHs3GjRtZvnw5gYGBzv1du3Zlzpw5HgvOZ7id3JxbLFNEREQKxq22pAULFjBnzhyuuOIKlxlzGzVqxM6dOz0WnM9wI7k5lJzBrqNpmE1whfrbiIiIFJhbNTdHjx6lcuXKefanpaVpeYD8uDHHzapdjiapy6tGEBHkXxxRiYiI+CS3kptWrVqxcOFC5+PchObtt9+mXbt2nonMV6SfgIxTjvsVEwp82prEk4BGSYmIiBSWW81SkydPpkePHmzZsoWcnBxefvlltmzZwq+//srPP//s6RjLttxam7BYCCj4DMN7jqcBcFlMWHFEJSIi4rPcqrm58sor2bhxIzk5OTRu3JjFixdTuXJlVq1aRcuWLT0dY9nmZmfivSfTAaheSUsuiIiIFEaha26ys7O5//77GTNmDG+99VZxxORb3JjjJsdm58CpDADitZ6UiIhIoRS65sbf358vvviiOGLxTW7U3BxMzsBmNwjwM1NZk/eJiIgUilvNUr1792bBggUeC2LGjBkkJCQQGBhI27ZtWbNmzQWPveqqqzCZTHm26667zmPxeNTJwo+U2nvC0SRVrUIQZrNGn4mIiBSGWx2K69aty8SJE1m5ciUtW7YkJCTE5fnhw4cXuKw5c+YwcuRIZs2aRdu2bZk+fTrdunVj27Zt+Q43nz9/PllZWc7Hx48fp2nTptx6663uvJTi50bNTW5/m2qRapISEREpLLeSm3feeYcKFSqwdu1a1q5d6/KcyWQqVHIzdepUBg8ezKBBgwCYNWsWCxcu5N133+XJJ5/Mc3xkZKTL488++4zg4ODSmdxkpDhWBAeoWPA+N3tPnAEgvmJQcUQlIiLi09xKbhITE533DcMAcGvyvqysLNauXcvo0aOd+8xmM127dmXVqlUFKuOdd97htttuy1N7lCszM5PMzEzn45SUlELH6bbcJqngKAgML/BpzpFSqrkREREpNLf63IAjqbj88ssJDAwkMDCQyy+/nLfffrtQZRw7dgybzUZMTIzL/piYGA4dOnTJ89esWcMff/zBvffee8FjpkyZQkREhHOLj48vVIxF4uYw8KSzfW7ildyIiIgUmlvJzdixY3n44Yfp1asXc+fOZe7cufTq1YtHHnmEsWPHejrGC3rnnXdo3Lgxbdq0ueAxo0ePJjk52bnt3bu3xOJze44bZ7OUkhsREZHCcqtZaubMmbz11lvcfvvtzn033HADTZo04aGHHmLixIkFKicqKgqLxcLhw4dd9h8+fJgqVapc9Ny0tDQ+++yzS17LarVitXppOLUbyc2ZLBvHUh3NaPGR6nMjIiJSWG7V3GRnZ9OqVas8+1u2bElOTk6BywkICKBly5YsXbrUuc9ut7N06dJLrlE1d+5cMjMzufPOOwseeEk7sdtxW4jkZt/Z/jZhVj8tmCkiIuIGt5Kbu+66i5kzZ+bZ/+abb9KvX79ClTVy5EjeeustPvjgA7Zu3cqQIUNIS0tzjp7q37+/S4fjXO+88w69e/emUqVSvLBkEYeBa4V1ERGRwnOrWQocycXixYu54oorAFi9ejVJSUn079+fkSNHOo+bOnXqRcvp27cvR48eZezYsRw6dIhmzZqxaNEiZyfjpKQkzGbXHGzbtm2sWLGCxYsXuxt+8ctKh9MHHPcLsfRCbn+b6mqSEhERcYtbyc0ff/xBixYtANi5cyfg6D8TFRXFH3/84TyuoDUPw4YNY9iwYfk+t3z58jz76tWr5xyCXmqd3O24DYyAoIoFPs05UkqdiUVERNziVnKzbNkyT8fhe85vkipE89JeDQMXEREpErfnuZFLcHcY+Mmzw8DVLCUiIuIWJTfFxY0FMw3DYJ+apURERIpEyU1xcaPmJvlMNqczHUPpqym5ERERcYuSm+KSm9y4sWBmdJiVoABLcUQlIiLi85TcFIecTEje57hfiJqbcyOl1N9GRETEXUpuisOpJDDs4B8CoZULfFruBH4aKSUiIuI+JTfFoajDwNXfRkRExG1KborDidyRUgXvbwMaBi4iIuIJSm6Kg5tz3OzTBH4iIiJFpuSmODiTm4LX3NjtBvtya27ULCUiIuI2JTfFwY2am8OnM8iy2bGYTcRGBBZTYCIiIr5PyY2n2XLg1B7H/UIkN7lz3MRVCMTPoo9FRETEXfor6mnJe8GeAxYrhMUV+DSNlBIREfEMJTeedvK8kVLmgr+9zjlulNyIiIgUiZIbT3N3NfCzzVLVKym5ERERKQolN552ovCrgcO5ZqlqWnpBRESkSJTceJpzwcyEQp2mpRdEREQ8Q8mNp7nRLJWZY+NQSgagPjciIiJFpeTGk+x2t5qlDpzKwDAgyN9CVGhAMQUnIiJSPii58aTTB8CWCWY/iIgv8Gnn97cxFWKhTREREclLyY0n5dbaVKgBFr8Cn5bb36a6+tuIiIgUmZIbT3JzGHiSFswUERHxGCU3nuTGgpkA+87OcaNh4CIiIkWn5MaT3J3AT8PARUREPEbJjScVcQI/DQMXEREpOiU3nmIYbtXcpGbmcDI9G4D4SDVLiYiIFJWSG09JOwrZaWAyQ4XqBT4tt9amYrA/YYH+xRWdiIhIuaHkxlNya20iqoGftcCnaaSUiIiIZxV8Mha5uNAYuOopsBSu9kX9bURERDxLyY2nRNaEq0YV+rR9J88OA1d/GxEREY9Qs5SXqeZGRETEs5TceJmWXhAREfEsJTdeZBgGe8/OTqwOxSIiIp6h5MaLjqVmcSbbhskEcRUCvR2OiIiIT1By40W5TVJVwgOx+lm8HI2IiIhvUHLjRepMLCIi4nlKbrxIw8BFREQ8T8mNF+XW3GiklIiIiOcoufGiJDVLiYiIeJySGy/K7VCsYeAiIiKeo+TGS3Jsdg6cygAgXn1uREREPEbJjZccTM7AZjcIsJiJCdMcNyIiIp6i5MZLcpukqlYMwmw2eTkaERER36Hkxkv2adkFERGRYqHkxkvOjZRSfxsRERFPUnLjJRopJSIiUjyU3HiJll4QEREpHkpuvGTvydw+N2qWEhER8SQlN16QkW3j6OlMQDU3IiIinqbkxgv2ne1vE2b1o0Kwv5ejERER8S1Kbrwgd6RUtchgTCbNcSMiIuJJSm68YG/uHDcaBi4iIuJxXk9uZsyYQUJCAoGBgbRt25Y1a9Zc9PhTp04xdOhQYmNjsVqtXHbZZXz33XclFK1nOEdKaRi4iIiIx/l58+Jz5sxh5MiRzJo1i7Zt2zJ9+nS6devGtm3bqFy5cp7js7KyuOaaa6hcuTLz5s2jatWq7NmzhwoVKpR88EXgnONGNTciIiIe59XkZurUqQwePJhBgwYBMGvWLBYuXMi7777Lk08+mef4d999lxMnTvDrr7/i7+/oiJuQkFCSIXvEXi29ICIiUmy81iyVlZXF2rVr6dq167lgzGa6du3KqlWr8j3n66+/pl27dgwdOpSYmBguv/xyJk+ejM1mu+B1MjMzSUlJcdm8yTAMZ7NUdSU3IiIiHue15ObYsWPYbDZiYmJc9sfExHDo0KF8z9m1axfz5s3DZrPx3XffMWbMGF566SWeffbZC15nypQpREREOLf4+HiPvo7CSj6TzenMHACqaY4bERERj/N6h+LCsNvtVK5cmTfffJOWLVvSt29fnn76aWbNmnXBc0aPHk1ycrJz27t3bwlGnFduk1RUqJWgAItXYxEREfFFXutzExUVhcVi4fDhwy77Dx8+TJUqVfI9JzY2Fn9/fyyWc0lBgwYNOHToEFlZWQQEBOQ5x2q1YrVaPRt8EZxbMFOdiUVERIqD12puAgICaNmyJUuXLnXus9vtLF26lHbt2uV7TocOHdixYwd2u9257++//yY2NjbfxKY00oKZIiIixcurzVIjR47krbfe4oMPPmDr1q0MGTKEtLQ05+ip/v37M3r0aOfxQ4YM4cSJEzz88MP8/fffLFy4kMmTJzN06FBvvYRCy625UWdiERGR4uHVoeB9+/bl6NGjjB07lkOHDtGsWTMWLVrk7GSclJSE2Xwu/4qPj+eHH37gkUceoUmTJlStWpWHH36YUaNGeeslFFrSCa0GLiIiUpxMhmEY3g6iJKWkpBAREUFycjLh4eElfv2rX1zOrmNpzL63Le3rRJX49UVERMqiwvz9LlOjpco6u91g30lN4CciIlKclNyUoCOnM8my2bGYTcRGBHo7HBEREZ+k5KYE5XYmjo0IxM+it15ERKQ46C9sCdKyCyIiIsVPyU0JStIcNyIiIsVOyU0J2qth4CIiIsVOyU0JOrf0gmpuREREiouSmxK072yzlFYDFxERKT5KbkpIVo6dgykZgJqlREREipOSmxJy4NQZDAMC/c1Eh5aeVcpFRER8jZKbEnL+SCmTyeTlaERERHyXkpsSos7EIiIiJUPJTQlxDgOvqP42IiIixUnJTQlRzY2IiEjJUHJTQjQMXEREpGQouSkhe086mqW0rpSIiEjxUnJTAlIzcziRlgVojhsREZHipuSmBOSuBl4h2J+wQH8vRyMiIuLblNyUgL1aDVxERKTEKLkpAbn9bdQkJSIiUvyU3JQAZ82NOhOLiIgUOyU3JWDfSTVLiYiIlBQlNyUgSTU3IiIiJUbJTTEzDENLL4iIiJQgJTfF7HhaFmeybZhMUFXJjYiISLFTclPMcjsTx4QFYvWzeDkaERER36fkpphp2QUREZGSpeSmmOXW3FTTHDciIiIlQslNMdPsxCIiIiVLyU0x23tSw8BFRERKkpKbYqZh4CIiIiVLyU0xstkNDpzKXVdKNTciIiIlQclNMTqYfIYcu0GAxUxMeKC3wxERESkXlNwUo9xlF6pWDMJiNnk5GhERkfJByU0x2ne2v0019bcREREpMX7eDsCXaaSUiIiDYRjk5ORgs9m8HYqUYv7+/lgsRZ/NX8lNMdIcNyIikJWVxcGDB0lPT/d2KFLKmUwmqlWrRmhoaJHKUXJTjHKXXojX7MQiUk7Z7XYSExOxWCzExcUREBCAyaQ+iJKXYRgcPXqUffv2Ubdu3SLV4Ci5KUa5NTdaV0pEyqusrCzsdjvx8fEEB+t3oVxcdHQ0u3fvJjs7u0jJjToUF5OMbBtHTmcCapYSETGb9edGLs1TtXr6aSsm+852Jg61+lEh2N/L0YiIiJQfSm6Kyd7zhoGrfVlERApi+fLlmEwmTp065dFjyxslN8VEw8BFRKSw2rdvz8GDB4mIiPDoseWNkptiomHgIiLlS1ZWVpHLCAgIoEqVKgWq8S/MseWNkptiktssVV3DwEVEyqSrrrqKYcOGMWzYMCIiIoiKimLMmDEYhgFAQkICkyZNon///oSHh3PfffcBsGLFCjp27EhQUBDx8fEMHz6ctLQ0Z7mZmZmMGjWK+Ph4rFYrderU4Z133gHyNjXt2bOHXr16UbFiRUJCQmjUqBHfffddvscCfPHFFzRq1Air1UpCQgIvvfSSy2tKSEhg8uTJ3H333YSFhVG9enXefPPN4noLvUbJTTHJXVdKzVIiIq4MwyA9K6fEt9ykpDA++OAD/Pz8WLNmDS+//DJTp07l7bffdj7/4osv0rRpU9avX8+YMWPYuXMn3bt3p0+fPmzatIk5c+awYsUKhg0b5jynf//+fPrpp7zyyits3bqVN95444KT1g0dOpTMzEx++eUXNm/ezH/+858LHrt27Vr+/e9/c9ttt7F582bGjx/PmDFjeP/9912Oe+mll2jVqhXr16/nwQcfZMiQIWzbtq3Q701ppnluion63IiI5O9Mto2GY38o8etumdiN4IDC/dmLj49n2rRpmEwm6tWrx+bNm5k2bRqDBw8G4Oqrr+bRRx91Hn/vvffSr18/RowYAUDdunV55ZVX6NSpEzNnziQpKYnPP/+cJUuW0LVrVwBq1ap1wesnJSXRp08fGjdufMljp06dSpcuXRgzZgwAl112GVu2bOGFF15g4MCBzuN69uzJgw8+CMCoUaOYNm0ay5Yto169eoV6b0oz1dwUg+T0bE5n5ABaNFNEpCy74oorXPq0tGvXju3btzvXyGrVqpXL8Rs3buT9998nNDTUuXXr1s05U/OGDRuwWCx06tSpQNcfPnw4zz77LB06dGDcuHFs2rTpgsdu3bqVDh06uOzr0KGDS7wATZo0cd43mUxUqVKFI0eOFCieskI1N8Ugt9YmKjSg0P8liIj4uiB/C1smdvPKdT0tJCTE5XFqair3338/w4cPz3Ns9erV2bFjR6HKv/fee+nWrRsLFy5k8eLFTJkyhZdeeomHHnrI7Zj9/V3nXjOZTNjtdrfLK430l7cY7FV/GxGRCzKZTGXmH7/Vq1e7PP7tt98uuu5RixYt2LJlC3Xq1Mn3+caNG2O32/n555+dzVKXEh8fzwMPPMADDzzA6NGjeeutt/JNbho0aMDKlStd9q1cuZLLLrvMIyttlyVqlioGzv42GgYuIlKmJSUlMXLkSLZt28ann37Kq6++ysMPP3zB40eNGsWvv/7KsGHD2LBhA9u3b+err75ydihOSEhgwIAB3H333SxYsIDExESWL1/O559/nm95I0aM4IcffiAxMZF169axbNkyGjRokO+xjz76KEuXLmXSpEn8/ffffPDBB7z22ms89thjRX8jypiykTqXMedGSqm/jYhIWda/f3/OnDlDmzZtsFgsPPzww84h3/lp0qQJP//8M08//TQdO3bEMAxq165N3759ncfMnDmTp556igcffJDjx49TvXp1nnrqqXzLs9lsDB06lH379hEeHk737t2ZNm1avse2aNGCzz//nLFjxzJp0iRiY2OZOHGiS2fi8sJkuDM2rgxLSUkhIiKC5ORkwsPDi+UaA95dw89/H+X/bm7MbW2qF8s1RETKgoyMDBITE6lZsyaBgYHeDqdQrrrqKpo1a8b06dO9HUq5cbGfl8L8/S4VzVIzZswgISGBwMBA2rZty5o1ay547Pvvv4/JZHLZStsXRsPARUREvMfryc2cOXMYOXIk48aNY926dTRt2pRu3bpddFhaeHg4Bw8edG579uwpwYgvzm432HfSMTux+tyIiIiUPK/3uZk6dSqDBw9m0KBBAMyaNYuFCxfy7rvv8uSTT+Z7Tu64/NLoaGomWTl2LGYTsRVKV42SiIgU3PLly70dgrjJqzU3WVlZrF271mU4nNlspmvXrqxateqC56WmplKjRg3i4+O58cYb+fPPP0si3ALJ7UwcGxGIv8XrFWMiIiLljlf/+h47dgybzUZMTIzL/piYGA4dOpTvOfXq1ePdd9/lq6++4uOPP8Zut9O+fXv27duX7/GZmZmkpKS4bMVJq4GLiIh4V5mrWmjXrh39+/enWbNmdOrUifnz5xMdHc0bb7yR7/FTpkwhIiLCucXHxxdrfLmrgWsYuIiIiHd4NbmJiorCYrFw+PBhl/2HDx8ucJ8af39/mjdvfsEprUePHk1ycrJz27t3b5HjvhhN4CciIuJdXk1uAgICaNmyJUuXLnXus9vtLF26lHbt2hWoDJvNxubNm4mNjc33eavVSnh4uMtWnLT0goiIiHd5fbTUyJEjGTBgAK1ataJNmzZMnz6dtLQ05+ip/v37U7VqVaZMmQLAxIkTueKKK6hTpw6nTp3ihRdeYM+ePdx7773efBlOzmHgSm5ERES8wut9bvr27cuLL77I2LFjadasGRs2bGDRokXOTsZJSUkcPHjQefzJkycZPHgwDRo0oGfPnqSkpPDrr7/SsGFDb70Ep6wcOweS1edGREQc60iV9tmNx48fT7NmzTx+rLdp+QUP2n0sjateXE6gv5mtE7tjMpk8Wr6ISFlTlpdfKKqEhARGjBjBiBEjvB3KBaWmppKZmUmlSpU8eqy7PLX8gtebpXxJbmfiahWDldiIiEixys7Oxt/fv0hlhIaGEhoa6vFjvc3rzVK+xDkMvKKapEREyrI333yTuLg47Ha7y/4bb7yRu+++m507d3LjjTcSExNDaGgorVu35scff3T7eiaTiZkzZ9KjRw+CgoKoVasW8+bNcz6/e/duTCYTc+bMoVOnTgQGBvLJJ58A8Pbbb9OgQQMCAwOpX78+r7/+ukvZ+/bt4/bbbycyMpKQkBBatWrF6tWrgbxNTcuXL6dNmzaEhIRQoUIFOnTo4Fzi6J/H2u12Jk6cSLVq1bBarTRr1oxFixbliXn+/Pl07tyZ4OBgmjZtetFJej1FyY0HacFMEZECMAzISiv5rRC9MG699VaOHz/OsmXLnPtOnDjBokWL6NevH6mpqfTs2ZOlS5eyfv16unfvTq9evUhKSnL7bRkzZgx9+vRh48aN9OvXj9tuu42tW7e6HPPkk0/y8MMPs3XrVrp168Ynn3zC2LFjee6559i6dSuTJ09mzJgxfPDBB4CjKalTp07s37+fr7/+mo0bN/LEE0/kSdoAcnJy6N27N506dWLTpk2sWrWK++6774ItES+//DIvvfQSL774Ips2baJbt27ccMMNbN++3eW4p59+mscee4wNGzZw2WWXcfvtt5OTk+P2+1QQapbyoNxh4NWV3IiIXFh2OkyOK/nrPnUAAkIKdGjFihXp0aMHs2fPpkuXLgDMmzePqKgoOnfujNlspmnTps7jJ02axJdffsnXX3/NsGHD3Arv1ltvdY78nTRpEkuWLOHVV191qYkZMWIEN998s/PxuHHjeOmll5z7atasyZYtW3jjjTcYMGAAs2fP5ujRo/z+++9ERkYCUKdOnXyvn5KSQnJyMtdffz21a9cGoEGDBheM98UXX2TUqFHcdtttAPznP/9h2bJlTJ8+nRkzZjiPe+yxx7juuusAmDBhAo0aNWLHjh3Ur1+/0O9RQanmxoNyk5tqmsBPRKTM69evH1988QWZmZkAfPLJJ9x2222YzWZSU1N57LHHaNCgARUqVCA0NJStW7cWqebmn/O7tWvXLk/NTatWrZz309LS2LlzJ/fcc4+zP0xoaCjPPvssO3fuBGDDhg00b97cmdhcTGRkJAMHDqRbt2706tWLl19+2WW08vlSUlI4cOAAHTp0cNnfoUOHPDE3adLEeT93TrojR45cMp6iUM2NB+09qWHgIiKX5B/sqEXxxnULoVevXhiGwcKFC2ndujX//e9/mTZtGuCojViyZAkvvvgiderUISgoiFtuuYWsrKziiNwpJORczVNqaioAb731Fm3btnU5zmKxABAUVLi/R++99x7Dhw9n0aJFzJkzh2eeeYYlS5ZwxRVXuB3z+Z2ec5u48msW8yQlNx6SlpnDiTTHD7X63IiIXITJVODmIW8KDAzk5ptv5pNPPmHHjh3Uq1ePFi1aALBy5UoGDhzITTfdBDgSjd27dxfper/99hv9+/d3edy8efMLHh8TE0NcXBy7du2iX79++R7TpEkT3n77bU6cOFGg2huA5s2b07x5c0aPHk27du2YPXt2nuQmPDycuLg4Vq5cSadOnZz7V65cSZs2bQp0neKk5MZDcjsTRwT5Ex5YtKF5IiJSOvTr14/rr7+eP//8kzvvvNO5v27dusyfP59evXphMpkYM2ZMkWsj5s6dS6tWrbjyyiv55JNPWLNmDe+8885Fz5kwYQLDhw8nIiKC7t27k5mZyf/+9z9OnjzJyJEjuf3225k8eTK9e/dmypQpxMbGsn79euLi4vI0gyUmJvLmm29yww03EBcXx7Zt29i+fbtLwnW+xx9/nHHjxlG7dm2aNWvGe++9x4YNG5yjuLxJyY2HnEzLJiLIX52JRUR8yNVXX01kZCTbtm3jjjvucO6fOnUqd999N+3btycqKopRo0aRkpJSpGtNmDCBzz77jAcffJDY2Fg+/fTTS86+f++99xIcHMwLL7zA448/TkhICI0bN3ZOHBgQEMDixYt59NFH6dmzJzk5OTRs2NClw2+u4OBg/vrrLz744AOOHz9ObGwsQ4cO5f7778/32sOHDyc5OZlHH32UI0eO0LBhQ77++mvq1q1bpPfBEzRDsYdlZNsI9Ld4vFwRkbKoPM9QXBgmk4kvv/yS3r17ezsUr/LUDMUaLeVhSmxERES8S8mNiIhIMfrkk09chmqfvzVq1Mjb4fkk9bkREREpRjfccEOeodq5codJl7MeIsVOyY2IiEgxCgsLIywszNthlCtqlhIRERGfouRGRESKnZpdpCA89XOi5EZERIpNbp+S9PR0L0ciZUHu8hW5y0e4S31uRESk2FgsFipUqOBcKDE4ONi5vpDI+ex2O0ePHiU4OBg/v6KlJ0puRESkWFWpUgUo/pWgpewzm81Ur169yAmwkhsRESlWJpOJ2NhYKleuTHZ2trfDkVIsICAAs7noPWaU3IiISImwWCxF7kshUhDqUCwiIiI+RcmNiIiI+BQlNyIiIuJTyl2fm9wJglJSUrwciYiIiBRU7t/tgkz0V+6Sm9OnTwMQHx/v5UhERESksE6fPk1ERMRFjzEZ5WxObLvdzoEDBwgLC/P4RFIpKSnEx8ezd+9ewsPDPVq2FJw+h9JBn0PpoM+hdNDnUHSGYXD69Gni4uIuOVy83NXcmM1mqlWrVqzXCA8P1w9vKaDPoXTQ51A66HMoHfQ5FM2lamxyqUOxiIiI+BQlNyIiIuJTlNx4kNVqZdy4cVitVm+HUq7pcygd9DmUDvocSgd9DiWr3HUoFhEREd+mmhsRERHxKUpuRERExKcouRERERGfouRGREREfIqSGw+ZMWMGCQkJBAYG0rZtW9asWePtkMqd8ePHYzKZXLb69et7Oyyf98svv9CrVy/i4uIwmUwsWLDA5XnDMBg7diyxsbEEBQXRtWtXtm/f7p1gfdilPoeBAwfm+X50797dO8H6qClTptC6dWvCwsKoXLkyvXv3Ztu2bS7HZGRkMHToUCpVqkRoaCh9+vTh8OHDXorYdym58YA5c+YwcuRIxo0bx7p162jatCndunXjyJEj3g6t3GnUqBEHDx50bitWrPB2SD4vLS2Npk2bMmPGjHyff/7553nllVeYNWsWq1evJiQkhG7dupGRkVHCkfq2S30OAN27d3f5fnz66aclGKHv+/nnnxk6dCi//fYbS5YsITs7m2uvvZa0tDTnMY888gjffPMNc+fO5eeff+bAgQPcfPPNXozaRxlSZG3atDGGDh3qfGyz2Yy4uDhjypQpXoyq/Bk3bpzRtGlTb4dRrgHGl19+6Xxst9uNKlWqGC+88IJz36lTpwyr1Wp8+umnXoiwfPjn52AYhjFgwADjxhtv9Eo85dWRI0cMwPj5558Nw3D87Pv7+xtz5851HrN161YDMFatWuWtMH2Sam6KKCsri7Vr19K1a1fnPrPZTNeuXVm1apUXIyuftm/fTlxcHLVq1aJfv34kJSV5O6RyLTExkUOHDrl8PyIiImjbtq2+H16wfPlyKleuTL169RgyZAjHjx/3dkg+LTk5GYDIyEgA1q5dS3Z2tsv3oX79+lSvXl3fBw9TclNEx44dw2azERMT47I/JiaGQ4cOeSmq8qlt27a8//77LFq0iJkzZ5KYmEjHjh05ffq0t0Mrt3K/A/p+eF/37t358MMPWbp0Kf/5z3/4+eef6dGjBzabzduh+SS73c6IESPo0KEDl19+OeD4PgQEBFChQgWXY/V98Lxytyq4+K4ePXo47zdp0oS2bdtSo0YNPv/8c+655x4vRibifbfddpvzfuPGjWnSpAm1a9dm+fLldOnSxYuR+aahQ4fyxx9/qN+fl6jmpoiioqKwWCx5ersfPnyYKlWqeCkqAahQoQKXXXYZO3bs8HYo5Vbud0Dfj9KnVq1aREVF6ftRDIYNG8a3337LsmXLqFatmnN/lSpVyMrK4tSpUy7H6/vgeUpuiiggIICWLVuydOlS5z673c7SpUtp166dFyOT1NRUdu7cSWxsrLdDKbdq1qxJlSpVXL4fKSkprF69Wt8PL9u3bx/Hjx/X98ODDMNg2LBhfPnll/z000/UrFnT5fmWLVvi7+/v8n3Ytm0bSUlJ+j54mJqlPGDkyJEMGDCAVq1a0aZNG6ZPn05aWhqDBg3ydmjlymOPPUavXr2oUaMGBw4cYNy4cVgsFm6//XZvh+bTUlNTXf77T0xMZMOGDURGRlK9enVGjBjBs88+S926dalZsyZjxowhLi6O3r17ey9oH3SxzyEyMpIJEybQp08fqlSpws6dO3niiSeoU6cO3bp182LUvmXo0KHMnj2br776irCwMGc/moiICIKCgoiIiOCee+5h5MiRREZGEh4ezkMPPUS7du244oorvBy9j/H2cC1f8eqrrxrVq1c3AgICjDZt2hi//fabt0Mqd/r27WvExsYaAQEBRtWqVY2+ffsaO3bs8HZYPm/ZsmUGkGcbMGCAYRiO4eBjxowxYmJiDKvVanTp0sXYtm2bd4P2QRf7HNLT041rr73WiI6ONvz9/Y0aNWoYgwcPNg4dOuTtsH1Kfu8/YLz33nvOY86cOWM8+OCDRsWKFY3g4GDjpptuMg4ePOi9oH2UyTAMo+RTKhEREZHioT43IiIi4lOU3IiIiIhPUXIjIiIiPkXJjYiIiPgUJTciIiLiU5TciIiIiE9RciMiIiI+RcmNiJRLJpOJBQsWeDsMESkGSm5EpMQNHDgQk8mUZ+vevbu3QxMRH6C1pUTEK7p37857773nss9qtXopGhHxJaq5ERGvsFqtVKlSxWWrWLEi4GgymjlzJj169CAoKIhatWoxb948l/M3b97M1VdfTVBQEJUqVeK+++4jNTXV5Zh3332XRo0aYbVaiY2NZdiwYS7PHzt2jJtuuong4GDq1q3L119/7Xzu5MmT9OvXj+joaIKCgqhbt26eZExESiclNyJSKo0ZM4Y+ffqwceNG+vXrx2233cbWrVsBSEtLo1u3blSsWJHff/+duXPn8uOPP7okLzNnzmTo0KHcd999bN68ma+//po6deq4XGPChAn8+9//ZtOmTfTs2ZN+/fpx4sQJ5/W3bNnC999/z9atW5k5cyZRUVEl9waIiPu8vXKniJQ/AwYMMCwWixESEuKyPffcc4ZhOFZXfuCBB1zOadu2rTFkyBDDMAzjzTffNCpWrGikpqY6n1+4cKFhNpudK13HxcUZTz/99AVjAIxnnnnG+Tg1NdUAjO+//94wDMPo1auXMWjQIM+8YBEpUepzIyJe0blzZ2bOnOmyLzIy0nm/Xbt2Ls+1a9eODRs2ALB161aaNm1KSEiI8/kOHTpgt9vZtm0bJpOJAwcO0KVLl4vG0KRJE+f9kJAQwsPDOXLkCABDhgyhT58+rFu3jmuvvZbevXvTvn17t16riJQsJTci4hUhISF5mok8JSgoqEDH+fv7uzw2mUzY7XYAevTowZ49e/juu+9YsmQJXbp0YejQobz44osej1dEPEt9bkSkVPrtt9/yPG7QoAEADRo0YOPGjaSlpTmfX7lyJWazmXr16hEWFkZCQgJLly4tUgzR0dEMGDCAjz/+mOnTp/Pmm28WqTwRKRmquRERr8jMzOTQoUMu+/z8/JyddufOnUurVq248sor+eSTT1izZg3vvPMOAP369WPcuHEMGDCA8ePHc/ToUR566CHuuusuYmJiABg/fjwPPPAAlStXpkePHpw+fZqVK1fy0EMPFSi+sWPH0rJlSxo1akRmZibffvutM7kSkdJNyY2IeMWiRYuIjY112VevXj3++usvwDGS6bPPPuPBBx8kNjaWTz/9lIYNGwIQHBzMDz/8wMMPP0zr1q0JDg6mT58+TJ061VnWgAEDyMjIYNq0aTz22GNERUVxyy23FDi+gIAARo8eze7duwkKCqJjx4589tlnHnjlIlLcTIZhGN4OQkTkfCaTiS+//JLevXt7OxQRKYPU50ZERER8ipIbERER8SnqcyMipY5ay0WkKFRzIyIiIj5FyY2IiIj4FCU3IiIi4lOU3IiIiIhPUXIjIiIiPkXJjYiIiPgUJTciIiLiU5TciIiIiE9RciMiIiI+5f8B2faoJOdEOccAAAAASUVORK5CYII=",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.plot(history.history['precision'], label='precision')\n",
        "plt.plot(history.history['val_precision'], label='Validation precision')\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel('precision')\n",
        "plt.legend(['precision', 'val_precision'])\n",
        "plt.title('Training and Validation precision')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8Go_m8nCYJJb",
        "outputId": "8687cfbd-5bc6-4965-acbd-855142dfb479"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.plot(history.history['recall'], label='recall')\n",
        "plt.plot(history.history['val_recall'], label='Validation recall')\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel('recall')\n",
        "plt.legend(['recall', 'val_recall'])\n",
        "plt.title('Training and Validation recall')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "RHeWoDNDQIDr",
        "outputId": "332b1dd0-51cf-4cd6-b80b-eb4474c19c79"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "22/22 [==============================] - 2s 6ms/step\n",
            "Confusion Matrix\n",
            "Accuracy for Test Data =  99.8529411764706\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1200x800 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      1.00      1.00       377\n",
            "           1       0.99      1.00      1.00       102\n",
            "           2       1.00      1.00      1.00       102\n",
            "           3       1.00      1.00      1.00        78\n",
            "           4       1.00      1.00      1.00        21\n",
            "\n",
            "    accuracy                           1.00       680\n",
            "   macro avg       1.00      1.00      1.00       680\n",
            "weighted avg       1.00      1.00      1.00       680\n",
            "\n",
            "\n",
            "Cohen Kappa Score\n",
            "0.9976808668101346\n",
            "\n",
            "Precision vs Recall Curve\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1200x800 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "ROC Curve\n",
            "ROC AUC score: 0.9995617376299872\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1200x800 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "def model_evaluate(model, target, testX, testy):\n",
        "  y_pred = model.predict(testX)\n",
        "  y_pred = np.argmax(y_pred, axis=1)\n",
        "  y_test = np.argmax(testy, axis=1)\n",
        "\n",
        "  # Confusion matrix\n",
        "  print('Confusion Matrix')\n",
        "\n",
        "  cm = confusion_matrix(y_test, y_pred)\n",
        "  # Calculate accuracy\n",
        "  from sklearn.metrics import accuracy_score, cohen_kappa_score, roc_curve, auc, roc_auc_score, precision_recall_curve\n",
        "  conf_acc = accuracy_score(y_test, y_pred) * 100\n",
        "  print(\"Accuracy for Test Data = \", conf_acc)\n",
        "\n",
        "  plt.figure(figsize = (12,8))\n",
        "  sns.heatmap(cm, annot=True, fmt='d')\n",
        "  plt.title(\"Confusion Matrix for Test Data\")\n",
        "  plt.xlabel('Predicted')\n",
        "  plt.ylabel('Expected')\n",
        "  plt.show()\n",
        "\n",
        "  # Classification Report\n",
        "  print('\\nClassification Report')\n",
        "  print(classification_report(y_test, y_pred))\n",
        "\n",
        "  # Cohen Kappa Score\n",
        "  print('\\nCohen Kappa Score')\n",
        "  print(cohen_kappa_score(y_test, y_pred))\n",
        "\n",
        "  # Precision vs Recall Curve\n",
        "  print('\\nPrecision vs Recall Curve')\n",
        "  fig, c_ax = plt.subplots(1,1, figsize = (12, 8))\n",
        "\n",
        "  y_test = tf.keras.utils.to_categorical(y_test)\n",
        "  y_pred = tf.keras.utils.to_categorical(y_pred)\n",
        "  n_classes = 2\n",
        "\n",
        "  precision = dict()\n",
        "  recall = dict()\n",
        "\n",
        "  for i in range(n_classes):\n",
        "    precision[i], recall[i], _ = precision_recall_curve(y_test[:, i],\n",
        "                                                          y_pred[:, i])\n",
        "    plt.plot(recall[i], precision[i], lw=2, label='class {}'.format(i))\n",
        "\n",
        "  plt.xlabel(\"recall\")\n",
        "  plt.ylabel(\"precision\")\n",
        "  plt.legend(loc=\"best\")\n",
        "  plt.title(\"Precision vs. Recall curve\")\n",
        "  plt.show()\n",
        "\n",
        "  # ROC Curve\n",
        "  print('\\nROC Curve')\n",
        "  average = 'macro'\n",
        "  fig, c_ax = plt.subplots(1,1, figsize = (12, 8))\n",
        "  for (idx, c_label) in enumerate(target):\n",
        "      fpr, tpr, thresholds = roc_curve(y_test[:,idx].astype(int), y_pred[:,idx])\n",
        "      c_ax.plot(fpr, tpr, label = '%s (AUC:%0.2f)'  % (c_label, auc(fpr, tpr)))\n",
        "  c_ax.plot(fpr, fpr, 'b-', label = 'Random Guessing')\n",
        "  print('ROC AUC score:', roc_auc_score(y_test, y_pred, average=average))\n",
        "\n",
        "  c_ax.legend()\n",
        "  c_ax.set_xlabel('False Positive Rate')\n",
        "  c_ax.set_ylabel('True Positive Rate')\n",
        "  plt.title('ROC Curve')\n",
        "  plt.show()\n",
        "\n",
        "model_evaluate(model, ['Relaxing', 'Coffee time', 'Early morning', 'Cleanup', 'Sandwich time'], X_test, y_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 81
        },
        "id": "7wls-W7EQM7l",
        "outputId": "edb874b3-1a4e-488c-bd03-3011da119236"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "\n",
              "  <div id=\"df-6eb6e090-7a28-4902-b4c8-f98775614643\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>loss</th>\n",
              "      <th>accuracy</th>\n",
              "      <th>precision</th>\n",
              "      <th>recall</th>\n",
              "      <th>true positive</th>\n",
              "      <th>true negative</th>\n",
              "      <th>false positive</th>\n",
              "      <th>false negative</th>\n",
              "      <th>area under the curve</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.00327</td>\n",
              "      <td>0.999276</td>\n",
              "      <td>0.999276</td>\n",
              "      <td>0.999276</td>\n",
              "      <td>2759.0</td>\n",
              "      <td>11042.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.999773</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6eb6e090-7a28-4902-b4c8-f98775614643')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-6eb6e090-7a28-4902-b4c8-f98775614643 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-6eb6e090-7a28-4902-b4c8-f98775614643');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "text/plain": [
              "      loss  accuracy  precision    recall  true positive  true negative  \\\n",
              "0  0.00327  0.999276   0.999276  0.999276         2759.0        11042.0   \n",
              "\n",
              "   false positive  false negative  area under the curve  \n",
              "0             2.0             2.0              0.999773  "
            ]
          },
          "execution_count": 23,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "pd.DataFrame([[history.history['loss'][-1], history.history['accuracy'][-1], history.history['precision'][-1], history.history['recall'][-1], history.history['tp'][-1], history.history['tn'][-1], history.history['fn'][-1], history.history['fp'][-1], history.history['auc'][-1]]],  columns=[\"loss\", \"accuracy\", \"precision\", \"recall\", \"true positive\", \"true negative\", \"false positive\", \"false negative\", \"area under the curve\"])"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "gpuType": "T4",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}