Neural Networks

From the basic perceptron to convolutional, recurrent, and sequence models — how neural networks are structured and how they learn.

In This Section

Perceptrons & Feedforward Networks

Neurons, layers, activations, and the universal approximation theorem.

Backpropagation — How Gradients Flow

The computational graph, chain rule in practice, and how weights are updated.

CNNs — Convolutional Networks

Filters, pooling, receptive fields, and why convolutions work for images.

RNNs, LSTMs & Sequence Models

Recurrent architectures for sequences, the vanishing gradient problem, and LSTM gating.

Encoder-Decoder & Seq2Seq

The seq2seq architecture, attention in encoder-decoder models, and T5/BART.