Title: Principles of Deep Learning Prerequisites: MATH 304, MATH 251, MATH 411, MATH 679, or equivalent; approval of instructor. Course Description: The course provides a rigorous introduction to the theory and practice of deep learning, including topics concerning approximation, generalization and optimization. It addresses the theory of universal approximation, (stochastic) gradient-based optimizers and statistical learning bounds, but also computational aspects (backpropagation, batch normalization, ...). Avg time dedicated per week (estimate): 6-8hrs