Deep statistical comparison for meta-heuristic stochastic optimization algorithms

DOI10.1007/978-3-030-96917-2MaRDI QIDQ2139909zbMATH OpenOpenAlexFDO

Authors Tome Eftimov, Peter Korošec

Publication date 19 May 2022

Published in Natural Computing Series (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/978-3-030-96917-2

zbMATH Keywords

benchmarkingmetaheuristicsmulti-objective optimizationsingle-objective optimizationstatistical comparisonstochastic optimization

Mathematics Subject Classification ID

Research exposition (monographs, survey articles) pertaining to operations research and mathematical programming (90-02) Stochastic programming (90C15) Approximation methods and heuristics in mathematical programming (90C59)

Review text

When a researcher presents a new algorithm, it is natural to compare it to the existing ones in order to prove that the new one is better. The authors of this book describe in detail the methodology called deep statistical comparison that allows to compare the metaheuristic stochastic optimization methods. The book consists of three parts. After a brief introduction in Chapter 1, in Chapters 2--4 the authors present the basic issues in optimization, benchmarking and statistical analysis. Chapters 5--7 present the deep statistical comparison applied to metaheuristic stochastic programming algorithms. Finally, in Chapter 8 the authors describe the online tool for deep statistical comparison. The book is well written and the presentation is easy to follow. It will be useful to students and researchers dealing with metaheuristic stochastic optimization, but also to practitioners who want to know how to choose the best methods to solve the real-life problems they face.

Recommendations

• A novel statistical approach for comparing meta-heuristic stochastic optimization algorithms according to the distribution of solutions in the search space
• Methods to compare expensive stochastic optimization algorithms with random restarts
• Theoretical framework for comparing several stochastic optimization approaches
• Stochastic global optimization.
• Metaheuristics for Hard Optimization

Cited in

(3)

• A novel statistical approach for comparing meta-heuristic stochastic optimization algorithms according to the distribution of solutions in the search space
• How to assess and report the performance of a stochastic algorithm on a benchmark problem: \textit{mean} or \textit{best} result on a number of runs?
• A statistical learning based approach for parameter fine-tuning of metaheuristics

This page was built for publication: Deep statistical comparison for meta-heuristic stochastic optimization algorithms

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2139909)