Generative Adversarial Model-Based Optimization via Source Critic Regularization

Michael S Yao, Yimeng Zeng, Hamsa Bastani, Jacob R. Gardner, James Gee, Osbert Bastani

Keywords: Offline Optimization, Bayesian Optimization, Surrogate Objectives

TL;DR: We show that source critic adversarial networks can effectively regularize offline optimizers for generative tasks in medicine and the sciences.

Abstract: Offline model-based optimization seeks to optimize against a learned surrogate model without querying the true oracle objective function during optimization. Such tasks are commonly encountered in protein design, robotics, and clinical medicine where evaluating the oracle function is prohibitively expensive. However, inaccurate surrogate model predictions are frequently encountered along offline optimization trajectories. To address this limitation, we propose *generative adversarial model-based optimization* using **adaptive source critic regularization (aSCR)**—a task- and optimizer- agnostic framework for constraining the optimization trajectory to regions of the design space where the surrogate function is reliable. We propose a computationally tractable algorithm to dynamically adjust the strength of this constraint, and show how leveraging aSCR with standard Bayesian optimization outperforms existing methods on a suite of offline generative design tasks. Our code is available at https://github.com/michael-s-yao/gabo.

Primary Area: Probabilistic methods (for example: variational inference, Gaussian processes)

Submission Number: 7166