Offline Model-Based Optimization via Policy-Guided Gradient Search

arXiv:2405.05349v1 Announce Type: new
Abstract: Offline optimization is an emerging problem in many experimental engineering domains including protein, drug or aircraft design, where online experimentation to collect evaluation data is too expensive or dangerous. To avoid that, one has to optimize an unknown function given only its offline evaluation at a fixed set of inputs. A naive solution to this problem is to learn a surrogate model of the unknown function and optimize this surrogate instead. However, such a naive …

abstract aircraft arxiv cs.ai cs.lg data design domains engineering evaluation experimental experimentation function gradient offline optimization policy protein search set type via