Nonlinear Data, Activation Functions, and Classification
The network models introduced in the first unit capture linear relationships, where a change in one variable results in a proportional change in another variable. However, the world is not always linear. This unit explores how to create network models that can handle nonlinear data. You will work with different kinds of activation functions, explore their properties, and see how adding them to a network model enables it to capture nonlinear data.
You will also apply neural networks to classification problems, where the goal is to identify to which class or cluster data belong. You will learn about linear vs nonlinear decision boundaries, loss functions for classification problems, and how to train a neural network to decode visual stimuli from neural spike recordings.