Fitness Agnostic Adaptive Sampling Lexicase Selection

Jared Moore; Adam Stanton

doi:10.1162/isal_a_00595

Fitness Agnostic Adaptive Sampling Lexicase Selection

Jared Moore, Adam Stanton

Research output: Chapter in Book/Published conference output › Conference publication

Abstract

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

Original language	English
Title of host publication	Proceedings of the 2023 Conference on Artificial Life
Number of pages	8
DOIs	https://doi.org/10.1162/isal_a_00595
Publication status	E-pub ahead of print - 24 Jul 2023
Event	ALIFE 2023: Ghost in the Machine: The 2023 Conference on Artificial Life - Sapporo, Japan Duration: 24 Jul 2023 → 28 Jul 2023 https://2023.alife.org/

Publication series

Name	Artificial Life Conference Proceedings
Publisher	MIT Press

Conference

Conference	ALIFE 2023: Ghost in the Machine
Abbreviated title	ALIFE 2023
Country/Territory	Japan
City	Sapporo
Period	24/07/23 → 28/07/23
Internet address	https://2023.alife.org/

Bibliographical note

© 2023 Massachusetts Institute of Technology Published under a Creative Commons Attribution 4.0 International (CC BY 4.0) license
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. For a full description of the license, please visit https://creativecommons.org/licenses/by/4.0/legalcode.

Access to Document

10.1162/isal_a_00595Licence: CC BY 4.0

Fitness Agnostic Adaptive Sampling Lexicase Selection
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. For a full description of the license, please visit https://creativecommons.org/licenses/by/4.0/legalcode.
Final published version, 692 KBLicence: CC BY 4.0

Cite this

@inproceedings{c1f9df6db73a4af790875128c3d0afe1,

title = "Fitness Agnostic Adaptive Sampling Lexicase Selection",

abstract = "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",

author = "Jared Moore and Adam Stanton",

note = "{\textcopyright} 2023 Massachusetts Institute of Technology Published under a Creative Commons Attribution 4.0 International (CC BY 4.0) license This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. For a full description of the license, please visit https://creativecommons.org/licenses/by/4.0/legalcode.; ALIFE 2023: Ghost in the Machine : The 2023 Conference on Artificial Life, ALIFE 2023 ; Conference date: 24-07-2023 Through 28-07-2023",

year = "2023",

month = jul,

day = "24",

doi = "10.1162/isal_a_00595",

language = "English",

series = "Artificial Life Conference Proceedings",

publisher = "MIT Press",

booktitle = "Proceedings of the 2023 Conference on Artificial Life",

url = "https://2023.alife.org/",

}

TY - GEN

T1 - Fitness Agnostic Adaptive Sampling Lexicase Selection

AU - Moore, Jared

AU - Stanton, Adam

N1 - © 2023 Massachusetts Institute of Technology Published under a Creative Commons Attribution 4.0 International (CC BY 4.0) license This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. For a full description of the license, please visit https://creativecommons.org/licenses/by/4.0/legalcode.

PY - 2023/7/24

Y1 - 2023/7/24

N2 - 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

AB - 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

UR - https://direct.mit.edu/isal/proceedings/isal/35/17/116898

U2 - 10.1162/isal_a_00595

DO - 10.1162/isal_a_00595

M3 - Conference publication

T3 - Artificial Life Conference Proceedings

BT - Proceedings of the 2023 Conference on Artificial Life

T2 - ALIFE 2023: Ghost in the Machine

Y2 - 24 July 2023 through 28 July 2023

ER -

Fitness Agnostic Adaptive Sampling Lexicase Selection

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this