Conservative Objective Models for Effective Offline Model-Based Optimization

In this paper, we aim to solve data-driven model-based optimization (MBO) problems, where the goal is to find a design input that maximizes an unknown objective function provided access to only a static dataset of inputs and their corresponding objective values. Such data-driven optimization procedures are the only practical methods in many real-world domains where active data collection is expensive (e.g., when optimizing over proteins) or dangerous (e.g., when optimizing over aircraft designs, actively evaluating malformed aircraft designs is unsafe). Typical methods for MBO that optimize the input against a learned model of the unknown score function are affected by erroneous overestimation in the learned model caused due to distributional shift, that drives the optimizer to low-scoring or invalid inputs. To overcome this, we propose conservative objective models (COMs), a method that learns a model of the objective function which lower bounds the actual value of the ground-truth objective on out-of-distribution inputs and uses it for optimization. In practice, COMs outperform a number existing methods on a wide range of MBO problems, including optimizing controller parameters, robot morphologies, and superconducting materials.