Deep Learning
Layers, neurons, trainable parameters, and how values flow.
How to write this in terms of matrix multiplication.
Finding the gradients of the loss with respect to the weights.