This course introduces the foundations and practical aspects of Deep Learning and Generative AI, combining theory with hands-on experience. It covers neural network models for computer vision and natural language processing, including CNNs, RNNs, LSTMs, and generative models such as autoencoders, VAEs, GANs, and large language models.

Co-requisite: CS2008 : Machine Learning Practice

Lectures-Tutorials-Practicals-Credits: 4-0-0-4

Week 1: Introduction to Deep Learning; Artificial neurons, neural networks, layers; Activation functions and loss metrics.

Week 2: Building simple neural networks using TensorFlow/Keras, Experimentation with activation functions and optimization methods.

Week 3: Introduction to Convolutional Neural Networks (CNNs), CNN architecture basics, convolution and pooling layers.

Week 4: Modelling Vision - CNN; CNN-based image classification (e.g., MNIST, CIFAR-10).

Week 5: Modelling Sequential Data - sequence models; Recurrent Neural Networks (RNNs), LSTMs.

Week 6: Sentiment analysis with LSTM; Text generation with simple RNN/LSTM models; Time series prediction tasks.

Week 7: Introduction to Generative AI, Autoencoders and Variational Autoencoders (VAEs) basics, GANs.

Week 8: Diffusion Probabilistic models, training and inference.

Week 9: GANs and Pretrained Diffusion Models to generate images

Week 10: Large Language Models - Transformer Architecture; Word Embeddings, Tokenization and attention.

Week 11: BERT like models for NLP tasks; Decoders for Text Generation; Machine Translation; Fine-tuning LLMs (PEFT, LoRA )

Week 12: Prompting Techniques, prompt fine-tuning and other methods