Robust and Practical Bayesian Optimization and Beyond

The central task in many interactive machine learning systems can be formalized as the sequential optimization of a black-box function. Bayesian optimization (BO) is a powerful model-based framework for adaptive experimentation, where the primary goal is the optimization of the black-box function via sequentially chosen decisions. In many real-world tasks, it is essential for the decisions to be robust against, e.g., adversarial failures and perturbations, dynamic and time-varying phenomena, a mismatch between simulations and reality, etc. Under such requirements, the standard methods and BO algorithms become inadequate. In this talk, we discuss algorithms with provable regret guarantees that can enhance robust and adaptive decision making in BO and related problems. We also consider associated robust submodular and non-submodular optimization problems, and present practical and efficient algorithms with improved robustness and constant factor approximation guarantees. Finally, we demonstrate the robust performance of our algorithms in numerous real-world applications (e.g., environmental monitoring and recommender systems) and tasks (e.g., robot pushing and feature selection). 

Key words: Bayesian optimization, Gaussian process, Submodularity, Robust optimization, Regret bounds, Level-set estimation, Non-submodular optimization

About the speaker — Volkan Cevher received the B.Sc. (valedictorian) in electrical engineering from Bilkent University in Ankara, Turkey, in 1999 and the Ph.D. in electrical and computer engineering from the Georgia Institute of Technology in Atlanta, GA in 2005. He was a Research Scientist with the University of Maryland, College Park from 2006-2007 and also with Rice University in Houston, TX, from 2008-2009. Currently, he is an Associate Professor at the Swiss Federal Institute of Technology Lausanne and a Faculty Fellow in the Electrical and Computer Engineering Department at Rice University. His research interests include signal processing theory, machine learning, convex optimization, and information theory. Dr. Cevher was the recipient of the IEEE Signal Processing Society Best Paper Award in 2016, a Best Paper Award at CAMSAP in 2015, a Best Paper Award at SPARS in 2009, and an ERC CG in 2016 as well as an ERC StG in 2011.