Build neural networks with tensors, layers, activation functions, backpropagation, optimization, regularization, and validation. You will train small models while watching for overfitting, data leakage, and compute limits.
