February 2019
Beginner to intermediate
308 pages
7h 42m
English
For the final output layer, we need an activation function that makes a prediction on the class of the label. For this project, we are making a simple binary prediction on the class: 1 for patients with onset of diabetes and 0 for patients without the onset of diabetes. The sigmoid activation function is ideal for binary classification problems.
Mathematically, we can represent the sigmoid activation function as follows:

Although this looks complicated, the underlying function is actually pretty simple. The Sigmoid Activation Function simply takes a value and squashes it between 0 and 1:
If the transformed value ...