Optimization of subsurface models with multiple criteria using Lexicase Selection

He, Yifan; Aranha, Claus; Hallam, Antony; Chassagne, Romain

doi:10.1016/j.orp.2022.100237

Cited by 5 publications

(2 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lexicase selection is a many-objective selection operator that has found success in a variety of domains from genetic programming (Spector, 2012), geosciences (He et al, 2022) and ER (Moore and Stanton, 2017). Lexicase is effective at maintaining population diversity and specialists within a population which contribute to its success in many-objective search problems (Helmuth et al, 2016;Dolson and Ofria, 2018).…”

Section: Background and Related Workmentioning

confidence: 99%

“…As individuals gain proficiency within the population, selection, recombination, and mutation further refine individuals to solve challenging tasks. Lexicase selection (Spector, 2012) has evolved solutions in many-objective optimization problems across a variety of problem domains (Helmuth et al, 2014;He et al, 2022). Applied to evolutionary robotics (ER) (Floreano et al, 2008;Doncieux et al, 2015), Lexicase can effectively leverage modal objectives (e.g.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fitness Agnostic Adaptive Sampling Lexicase Selection

Moore,

Stanton

2023

The 2023 Conference on Artificial Life

View full text Add to dashboard Cite

Lexicase selection is an effective many-objective evolutionary algorithm across many problem domains. Lexicase can be computationally expensive, especially in areas like evolutionary robotics where individual objectives might require their own physics simulation. Improving the efficiency of Lexicase selection can reduce the total number of evaluations thereby lowering computational overhead. Here, we introduce a fitness agnostic adaptive objective sampling algorithm using the filtering efficacy of objectives to adjust their frequency of occurrence as a selector. In a set of binary genome maximization tasks modeled to emulate evolutionary robotics situations, we show that performance can be maintained while computational efficiency increases as compared to ϵ-Lexicase.

show abstract