BOTORCH: A Framework for Efficient Monte-Carlo Bayesian Optimization
Max Balandat, Brian Karrer, Daniel Jiang, Sam Daulton, Ben Letham, Andrew Gordon Wilson, Eytan Bakshy
I lead the Modeling & Optimization team within the Adaptive Experimentation group on Facebook’s Core Data Science team. We focus on developing methods and tools for probabilistic modeling and sample-efficient optimization, and apply them to a broad range of applications across the company, including infrastructure optimization, AutoML, online A/B tests, ranking systems, and AR/VR. I also lead the development of BoTorch, an open-source library for Bayesian Optimization in PyTorch.
In the past, I have also worked on the intersection of Machine Learning and Econometrics, in particular on how to utilize Machine Learning algorithms to perform causal inference in experimental and non-experimental settings.
I hold an MA in Mathematics and a PhD in Electrical Engineering and Computer Sciences from UC Berkeley.
Interests
Bayesian Optimization, Gaussian processes, probabilistic modeling, causal inference and machine Learning
Latest Publications
NeurIPS
Efficient Nonmyopic Bayesian Optimization via One-Shot Multi-Step Trees
Shali Jiang, Daniel Jiang, Max Balandat, Brian Karrer, Jacob R. Gardner, Roman Garnett
