Kidney Disease Tumor or Normal Classification

Project Overview

This project aims to develop a machine learning model to classify kidney images as either tumor or normal. Early and accurate classification is crucial for the effective treatment of kidney diseases.

Installation

To run this project locally, follow these steps:

1. Clone the repository:

   git clone https://github.com/AbQaadir/Kidney-Disease-Classification.git
   cd kidney-tumor-classification

2. Create a virtual environment and activate it:

   python -m venv venv
   source venv/bin/activate   # On Windows use `venv\Scripts\activate`

3. Install the required packages:

   pip install -r requirements.txt

Running the Pipeline

Individual Pipeline Steps

1. Data Ingestion:

   python src/KDC/pipeline/step_01_data_ingestion.py

2. Prepare Base Model:

   python src/KDC/pipeline/step_02_prepare_base_model.py

3. Model Training:

   python src/KDC/pipeline/step_03_model_training.py

4. Model Evaluation:

   python src/KDC/pipeline/step_04_model_evaluation.py

Running the Complete Pipeline

To run the entire pipeline from data preprocessing to model evaluation, use the following command:

python main.py

Experiment Management

We use Data Version Control (DVC) for experiment management. To reproduce experiments, run:

dvc repro

Experiment Tracking

For detailed experiment tracking, See the DagsHub Repo (mlflow)

https://dagshub.com/AbQaadir/Kidney-Disease-Classification

Additional Notes

• Ensure that you have all necessary data files in the correct directories before running the pipelines.

• If you encounter any issues, please refer to the logs/ directory for detailed logs of each pipeline step.