Deep Learning

Catalog description

This course provides a comprehensive exploration of the theoretical foundations and current techniques in deep learning. Moving beyond introductory machine learning, the curriculum delves into the core principles of learning hierarchical representations from complex data. Topics include advanced optimization and regularization techniques, convolutional and recurrent neural network architectures, the Transformer architecture and attention mechanisms, and modern learning paradigms such as generative modeling (VAEs, GANs, Diffusion Models), and graph neural networks (GNNs). The course emphasizes both the mathematical underpinnings of these models and their practical implementation for solving real-world problems in computer vision, natural language processing, and other scientific domains.