BAYESIAN OPTIMIZATION AND STATISTICAL REINFORCEMENT LEARNING

Credit Hours: 3 · Graduate

Bayesian optimization is a powerful machine learning technique for optimizing black-box functions with expensive evaluations. This course provides an in-depth introduction to Bayesian optimization, covering fundamental principles, mathematical foundations, and real-world applications. Students will explore experimental design, Gaussian processes, nonparametric methods, surrogate modeling, reinforcement learning, and bandit algorithms. The course will emphasize mathematical theoretical understanding in practical implementation, preparing students for research-level statistical reinforcement learning. Instructor Permission Required. Recommended Prerequisite(s): STAT 532 /533 or STAT 581 /582