On the Convergence of Prior-Guided Zeroth-Order Optimization Algorithms

Abstract

Zeroth-order (ZO) optimization is widely used to handle challenging tasks, such as query-based black-box adversarial attacks and reinforcement learning. Various attempts have been made to integrate prior information into the gradient estimation procedure based on finite differences, with promising empirical results. However, their convergence properties are not well understood. This paper makes an attempt to fill up this gap by analyzing the convergence of prior-guided ZO algorithms under a greedy descent framework with various gradient estimators. We provide a convergence guarantee for the prior-guided random gradient-free (PRGF) algorithms. Moreover, to further accelerate over greedy descent methods, we present a new accelerated random search (ARS) algorithm that incorporates prior information, together with a convergence analysis. Finally, our theoretical results are confirmed by experiments on several numerical benchmarks as well as adversarial attacks.

arXiv commentNeurIPS 2021; code available at https://github.com/csy530216/pg-zoo

References in corpus (10)

• Practical Bayesian Optimization of Machine Learning Algorithms
• ZOO: Zeroth Order Optimization based Black-box Attacks to Deep Neural Networks without Training Substitute Models
• Evolution Strategies as a Scalable Alternative to Reinforcement Learning
• Learning to learn by gradient descent by gradient descent
• Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors
• Improving Black-box Adversarial Attacks with a Transfer-based Prior
• Guessing Smart: Biased Sampling for Efficient Black-Box Adversarial Attacks
• Gradientless Descent: High-Dimensional Zeroth-Order Optimization
• Guided evolutionary strategies: Augmenting random search with surrogate gradients
• Improving Gradient Estimation in Evolutionary Strategies With Past Descent Directions