Geometric Foundations of Machine Learning

3 Hours

Explore geometric foundations of modern deep and reinforcement learning. Examine the mathematical, statistical, and computational optimization fundamentals that are the basis of computational, data driven machine learning models and Markov decision making processes. Discuss how data efficient and continuous action spaces are harnessed to learn the free energy Hamiltonian underlying dynamical systems, and multi-player games. Only one of the following may be counted: Computer Science 363G, 378 (Topic: Geometric Fndtns Of ML), 378 (Topic: Geometric Fndtns Of Data Sci). The equivalent of three lecture hours a week for one semester.