Book 6 — Deep Learning

Neural networks built from the ground up.

Foundations

• Perceptron & neuron
• Layers, weights, bias, activation
• Forward propagation
• Loss & backpropagation
• Gradient descent

Architecture Components

• MLP
• ReLU / Sigmoid / Tanh / Softmax
• Batch normalization
• Dropout
• Weight initialization

Optimizers

• SGD & momentum
• AdaGrad / RMSProp
• Adam / AdamW

Status

Complete deep-dive coverage of neural network foundations, modern architecture components and regularization, and deep learning optimizers.