copy-edit: change 'GMM model'

Author: jakevdp

notebooks/05.12-Gaussian-Mixtures.ipynb

@@ -374,7 +374,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "With this in place, we can take a look at what the GMM model gives us for our initial data:"
+    "With this in place, we can take a look at what the four-component GMM gives us for our initial data:"
    ]
   },
   {
@@ -561,8 +561,8 @@
    "metadata": {},
    "source": [
     "Here the mixture of 16 Gaussians serves not to find separated clusters of data, but rather to model the overall *distribution* of the input data.\n",
-    "This is a generative model of the distribution, meaning that the GMM model gives us the recipe to generate new random data distributed similarly to our input.\n",
-    "For example, here are 400 new points drawn from this 16-component GMM model to our original data:"
+    "This is a generative model of the distribution, meaning that the GMM gives us the recipe to generate new random data distributed similarly to our input.\n",
+    "For example, here are 400 new points drawn from this 16-component GMM fit to our original data:"
    ]
   },
   {
@@ -604,7 +604,7 @@
     "The fact that GMM is a generative model gives us a natural means of determining the optimal number of components for a given dataset.\n",
     "A generative model is inherently a probability distribution for the dataset, and so we can simply evaluate the *likelihood* of the data under the model, using cross-validation to avoid over-fitting.\n",
     "Another means of correcting for over-fitting is to adjust the model likelihoods using some analytic criterion such as the [Akaike information criterion (AIC)](https://en.wikipedia.org/wiki/Akaike_information_criterion) or the [Bayesian information criterion (BIC)](https://en.wikipedia.org/wiki/Bayesian_information_criterion).\n",
-    "Scikit-Learn's ``GMM`` model actually includes built-in methods that compute both of these, and so it is very easy to operate on this approach.\n",
+    "Scikit-Learn's ``GMM`` estimator actually includes built-in methods that compute both of these, and so it is very easy to operate on this approach.\n",
     "\n",
     "Let's look at the AIC and BIC as a function as the number of GMM components for our moon dataset:"
    ]
@@ -725,8 +725,8 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "We have nearly 1,800 digits in 64 dimensions, and we can build a GMM model on top of these to generate more.\n",
-    "GMM can have difficulty converging in such a high dimensional space, so we will start with an invertible dimensionality reduction algorithm on the data.\n",
+    "We have nearly 1,800 digits in 64 dimensions, and we can build a GMM on top of these to generate more.\n",
+    "GMMs can have difficulty converging in such a high dimensional space, so we will start with an invertible dimensionality reduction algorithm on the data.\n",
     "Here we will use a straightforward PCA, asking it to preserve 99% of the variance in the projected data:"
    ]
   },