Deep Learning

Catalog description

This course provides a comprehensive and rigorous introduction to the field of Deep Learning, the engine behind modern breakthroughs in computer vision, natural language processing, and generative AI. Doctoral students will move beyond traditional machine learning to explore the design, training, and optimization of multi-layered artificial neural networks. The course begins with the mathematical foundations of computational graphs and backpropagation before advancing to specialized architectures. Students will conduct deep dives into Convolutional Neural Networks (CNNs) for spatial data and Recurrent Neural Networks (RNNs) for sequential data. Significant emphasis is placed on the Transformer architecture, which has revolutionized the field. Through hands-on research and implementation using frameworks, students will address complex challenges such as vanishing gradients, hyperparameter optimization, and the scaling laws of large-scale models.