Improving sampling in evolution strategies through mixture-based distributions built from past problem instances

Friess，Stephen1; Tiňo，Peter1; Menzel，Stefan2; Sendhoff，Bernhard2; Yao，Xin1,3

10.1007/978-3-030-58112-1_40

2020

0302-9743

1611-3349

LECT NOTES ARTIF INT

12269 LNCS

583-596

The notion of learning from different problem instances, although an old and known one, has in recent years regained popularity within the optimization community. Notable endeavors have been drawing inspiration from machine learning methods as a means for algorithm selection and solution transfer. However, surprisingly approaches which are centered around internal sampling models have not been revisited. Even though notable algorithms have been established in the last decades. In this work, we progress along this direction by investigating a method that allows us to learn an evolutionary search strategy reflecting rough characteristics of a fitness landscape. This latter model of a search strategy is represented through a flexible mixture-based distribution, which can subsequently be transferred and adapted for similar problems of interest. We validate this approach in two series of experiments in which we first demonstrate the efficacy of the recovered distributions and subsequently investigate the transfer with a systematic from the literature to generate benchmarking scenarios.

Algorithm configuration Continuous optimisation Evolution strategies Model-based optimisation Transfer learning

其他

英语

EI

20203909228666

Machine learning ; Learning algorithms ; Optimization ; Evolutionary algorithms

Artificial Intelligence:723.4 ; Machine Learning:723.4.2 ; Optimization Techniques:921.5

2-s2.0-85091287808

Scopus

1.CERCIA,School of Computer Science,University of Birmingham,Birmingham,United Kingdom
2.Honda Research Institute Europe,Offenbach a.M.,Carl-Legien-Str. 30,63073,Germany
3.Department of Computer Science and Engineering,Southern University of Science and Technology,Shenzhen,China

Friess，Stephen,Tiňo，Peter,Menzel，Stefan,et al. Improving sampling in evolution strategies through mixture-based distributions built from past problem instances[C],2020:583-596.