We study the fundamentals of Bayesian optimization and develop efficient Bayesian optimization methods for global optimization of expensive black-box functions originated from a range of different applications.

The optimization of an unknown function that is expensive to evaluate is an important problem in many areas of science and engineering. Bayesian optimization uses probabilistic methods to address this problem. Motivated by real-world applications in high-dimensional parameter-tuning for complex machine learning algorithms and expensive optimization problems/active learning problems in robotics, we study the theoretical understandings of Bayesian optimization, connections among existing methods, and develop efficient and provably correct global optimization methods for these applications.

Research Areas