Binary classification

Binary classification refers to problems with only two distinct classes. As we did in the previous chapter, we will generate a dataset using the convenience function, make_classification(), in the SciKit Learn library:

X, y = skds.make_classification(n_samples=200,
   n_features=2,
   n_informative=2, 
   n_redundant=0, 
   n_repeated=0,
   n_classes=2,
   n_clusters_per_class=1)
if (y.ndim == 1):
    y = y.reshape(-1,1)

The arguments to make_classification() are self-explanatory; n_samples is the number of data points to generate, n_features is the number of features to be generated, and n_classes is the number of classes, which is 2:

• n_samples is the number of data points to generate. We have kept it to 200 to keep the dataset small. 
• n_features is the number of features to be generated; we are using only two features so that we can keep it a simple problem to understand the TensorFlow commands.
• n_classes is the number of classes, which is 2 as it is a binary classification...