Part 2 - Deep Learning

Deep learning extends classical machine learning with neural networks that can learn directly from raw data — images, text, audio — without manual feature engineering. In this part we cover the fundamentals of neural networks using PyTorch and then apply them to practical problems.

We start with the basics of deep learning: network architectures, training, and evaluation. We then explore natural language processing with deep learning, including text classification and working with pre-trained transformer models. The remaining chapters provide overviews of image generation, reinforcement learning, and recommendation systems.

When you have finished reading this section and want to learn more about specific deep learning architectures, I recommend this curated list of explained papers: ML-Papers-Explained. This ML Papers directory lists papers in chronological order and also includes LLM papers you might find useful after finishing this book.