Week 7 covers the second pillar of Domain 3 (Modeling): deep learning. Unstructured data (images, text, audio) and complex nonlinear relationships are deep learning's domain. MLS-C01 doesn't probe framework-level coding but presumes conceptual understanding of perceptron, activation functions, loss, backpropagation, hyperparameters. Today we nail the big picture of how neural nets turn input into prediction, and how they learn.
A perceptron takes weighted input, adds bias, and passes through activation function.
z = w1*x1 + w2*x2 + ... + wn*xn + b
a = f(z) # f = activation function