A one-bit, comparison-based gradient estimator

DOI10.1016/J.ACHA.2022.03.003OpenAlexW3122422392MaRDI QIDQ2155805FDOQ2155805

Authors: HanQin Cai, Daniel McKenzie, Wotao Yin, Zhen-Liang Zhang 

Publication date: 15 July 2022

Published in: Applied and Computational Harmonic Analysis (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2010.02479

Recommendations

• Small errors in random zeroth-order optimization are imaginary
• Minimax efficient finite-difference stochastic gradient estimators using black-box function evaluations
• Black-box reductions for zeroth-order gradient algorithms to achieve lower query complexity
• Zeroth-order regularized optimization (ZORO): approximately sparse gradients and adaptive sampling

zbMATH Keywords

reinforcement learninghyperparameter tuningzeroth-order optimizationone-bit compressed sensingcomparison-based optimizationnormalized gradient descent

Mathematics Subject Classification ID

Computer science (68-XX) Information and communication theory, circuits (94-XX)

Cites Work

• Title not available (Why is that?)
• Bayesian optimization in a billion dimensions via random embeddings
• Title not available (Why is that?)
• CoSaMP: Iterative signal recovery from incomplete and inaccurate samples
• Adaptive stochastic approximation by the simultaneous perturbation method
• Parallel distributed block coordinate descent methods based on pairwise comparison oracle
• Preference-based reinforcement learning: a formal framework and a policy iteration algorithm
• Minimax number of strata for online stratified sampling given noisy samples
• Active subspaces. Emerging ideas for dimension reduction in parameter studies
• Restricted strong convexity and its applications to convergence analysis of gradient-type methods in convex optimization
• A dimensionality reduction technique for unconstrained global optimization of functions with low effective dimensionality
• Robust 1-bit Compressed Sensing and Sparse Logistic Regression: A Convex Programming Approach
• Derivative-free optimization methods
• Title not available (Why is that?)
• Zeroth-order regularized optimization (ZORO): approximately sparse gradients and adaptive sampling

Cited In (4)

• Global solutions to nonconvex problems by evolution of Hamilton-Jacobi PDEs
• A One-bit, Comparison-Based Gradient Estimator
• Small errors in random zeroth-order optimization are imaginary
• Stochastic zeroth order descent with structured directions

This page was built for publication: A one-bit, comparison-based gradient estimator

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2155805)