Improving exploration strategies in large dimensions and rate of convergence of global random search algorithms

DOI10.1007/S10898-023-01308-6arXiv2208.11542OpenAlexW4385225555MaRDI QIDQ6183083FDOQ6183083

Authors: Jack Noonan, A. Zhigljavsky 

Publication date: 26 January 2024

Published in: Journal of Global Optimization (Search for Journal in Brave)

Abstract: We consider global optimization problems, where the feasible region $X$ is a compact subset of $mathbb{R}^d$ with $dgeq10$. For these problems, we demonstrate the following. First: the actual convergence of global random search algorithms is much slower than that given by the classical estimates, based on the asymptotic properties of random points. Second: the usually recommended space exploration schemes are inefficient in the non-asymptotic regime. Specifically, (a) uniform sampling on entire~$X$ is much less efficient than uniform sampling on a suitable subset of $X$, and (b) the effect of replacement of random points by low-discrepancy sequences is negligible.

Full work available at URL: https://arxiv.org/abs/2208.11542

Recommendations

• Performance of global random search algorithms for large dimensions
• scientific article; zbMATH DE number 4020863
• Improving hit-and-run for global optimization
• Bayesian and High-Dimensional Global Optimization
• On the asymptotic tractability of global optimization

zbMATH Keywords

central limit theoremglobal optimizationstochastic optimizationspace-fillinglarge dimensionglobal random search

Mathematics Subject Classification ID

Mathematical programming (90Cxx)

Cites Work

• The design and analysis of computer experiments.
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Kernel techniques: From machine learning to meshless methods
• Random coverings in several dimensions
• A space quantization method for numerical integration
• Asymptotic quantization error of continuous signals and the quantization dimension
• On the worst-case optimal multi-objective global optimization
• Centroidal Voronoi Tessellations: Applications and Algorithms
• Title not available (Why is that?)
• Convergence properties of stochastic optimization procedures
• Minimization by Random Search Techniques
• Stochastic global optimization.
• Scattered Data Approximation
• Bayesian and High-Dimensional Global Optimization
• Minimax models in the theory of numerical methods. Transl. from the 1989 Russian orig. by Olga Chuyan
• Performance of global random search algorithms for large dimensions
• Discrete energy on rectifiable sets
• Title not available (Why is that?)
• Title not available (Why is that?)
• Pure random search with virtual extension of feasible region
• Bayesian quadrature, energy minimization, and space-filling design
• Efficient quantisation and weak covering of high dimensional cubes
• Covering of high-dimensional cubes and quantization
• Non-lattice Covering and Quantization of High Dimensional Sets
• Random Euclidean coverage from within

This page was built for publication: Improving exploration strategies in large dimensions and rate of convergence of global random search algorithms

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6183083)