Timezone: »

Black-Box Optimization with Local Generative Surrogates
Sergey Shirobokov · Vladislav Belavin · Michael Kagan · Andrei Ustyuzhanin · Atilim Gunes Baydin

Wed Dec 09 09:00 AM -- 11:00 AM (PST) @ Poster Session 3 #1055

We propose a novel method for gradient-based optimization of black-box simulators using differentiable local surrogate models. In fields such as physics and engineering, many processes are modeled with non-differentiable simulators with intractable likelihoods. Optimization of these forward models is particularly challenging, especially when the simulator is stochastic. To address such cases, we introduce the use of deep generative models to iteratively approximate the simulator in local neighborhoods of the parameter space. We demonstrate that these local surrogates can be used to approximate the gradient of the simulator, and thus enable gradient-based optimization of simulator parameters. In cases where the dependence of the simulator on the parameter space is constrained to a low dimensional submanifold, we observe that our method attains minima faster than baseline methods, including Bayesian optimization, numerical optimization and approaches using score function gradient estimators.

Author Information

Sergey Shirobokov (Imperial College London)
Vladislav Belavin (National Research University Higher School of Economics)
Michael Kagan (SLAC / Stanford)
Andrei Ustyuzhanin (National Research University Higher School of Economics)
Atilim Gunes Baydin (University of Oxford)

More from the Same Authors