Deep Learning: Master Neural Networks and AI

An introduction to deep learning, focusing on feed-forward deep neural networks and backpropagation. The module covers the basics of neural networks and how to optimize them.

This module covers Convolutional Neural Networks (CNNs), including important layers such as convolution and pooling. It also discusses various CNN applications.

This module provides practical deep learning tips including activation function selection, adaptive gradient descent learning methods, regularization, and dropout.

This module focuses on Recurrent Neural Networks (RNNs) used for sequential data, covering Vanilla RNNs and Long Short-term Memory (LSTM). It also discusses learning problems in RNNs.

This module covers generative models, particularly Generative Adversarial Networks (GANs) and Diffusion Models (DMs). It explains how these models work and their applications.

This module discusses the transformer, a powerful deep learning model used in natural language processing, computer vision, and spatio-temporal modelling. It covers self-attention mechanisms and transformer architecture.

This module covers neural network compression techniques, including network pruning, knowledge distillation, parameter quantization, and architecture design. It emphasizes the importance of model size reduction for deployment on resource-constrained devices.

This module discusses transfer learning techniques, including fine-tuning, multitask learning, domain adversarial training, and zero-shot learning. It explains how to reuse pre-trained models for new tasks.

This module contains the summative course assessment designed to evaluate understanding of the course material and assess the ability to apply the knowledge acquired throughout the course.