Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers

Pedro Carvalho; Bruno Ribeiro; Nuno M. Rodrigues; João E. Batista; Leonardo Vanneschi; Sara Silva

doi:10.1007/978-3-031-30229-9_42

Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers

Pedro Carvalho, Bruno Ribeiro, Nuno M. Rodrigues, João E. Batista, Leonardo Vanneschi, Sara Silva

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

30 Downloads (Pure)

Abstract

Feature selection is becoming an essential part of machine learning pipelines, including the ones generated by recent AutoML tools. In case of datasets with epistatic interactions between the features, like many datasets from the bioinformatics domain, feature selection may even become crucial. A recent method called SLUG has outperformed the state-of-the-art algorithms for feature selection on a large set of epistatic noisy datasets. SLUG uses genetic programming (GP) as a classifier (learner), nested inside a genetic algorithm (GA) that performs feature selection (wrapper). In this work, we pair GA with different learners, in an attempt to match the results of SLUG with less computational effort. We also propose a new feedback mechanism between the learner and the wrapper to improve the convergence towards the key features. Although we do not match the results of SLUG, we demonstrate the positive effect of the feedback mechanism, motivating additional research in this area to further improve SLUG and other existing feature selection methods.

Original language	English
Title of host publication	Applications of Evolutionary Computation
Subtitle of host publication	26th European Conference, EvoApplications 2023 Held as Part of EvoStar 2023 Brno, Czech Republic, April 12–14, 2023 Proceedings
Editors	João Correia, Stephen Smith, Raneem Qaddoura
Place of Publication	Gewerbestrasse
Publisher	Springer
Chapter	42
Pages	656-671
Number of pages	16
ISBN (Electronic)	978-3-031-30229-9
ISBN (Print)	978-3-031-30228-2
DOIs	https://doi.org/10.1007/978-3-031-30229-9_42
Publication status	Published - Apr 2023
Event	26th International Conference on the Applications of Evolutionary Computation - Brno, Brno, Czech Republic Duration: 12 Apr 2023 → 14 Apr 2023 Conference number: 26 https://www.evostar.org/2023/evoapps/

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	13989
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	26th International Conference on the Applications of Evolutionary Computation
Abbreviated title	EvoApplications 2023
Country/Territory	Czech Republic
City	Brno
Period	12/04/23 → 14/04/23
Internet address	https://www.evostar.org/2023/evoapps/

Keywords

Feature Selection
Epistasis
Genetic Algorithms
Genetic Programming
Decision Trees
Machine Learning
Genome-Wide Association Studies

Access to Document

10.1007/978-3-031-30229-9_42

Problems_Genetic_Algorithms_Nested_Classifiers_Final_versionAccepted author manuscript, 652 KBLicence: CC BY-NC

https://link.springer.com/10.1007/978-3-031-30229-9_42Licence: Unspecified

Cite this

Carvalho, P., Ribeiro, B., Rodrigues, N. M., Batista, J. E., Vanneschi, L., & Silva, S. (2023). Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers. In J. Correia, S. Smith, & R. Qaddoura (Eds.), Applications of Evolutionary Computation: 26th European Conference, EvoApplications 2023 Held as Part of EvoStar 2023 Brno, Czech Republic, April 12–14, 2023 Proceedings (pp. 656-671). (Lecture Notes in Computer Science; Vol. 13989). Springer. https://doi.org/10.1007/978-3-031-30229-9_42

Carvalho, Pedro ; Ribeiro, Bruno ; Rodrigues, Nuno M. et al. / Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers. Applications of Evolutionary Computation: 26th European Conference, EvoApplications 2023 Held as Part of EvoStar 2023 Brno, Czech Republic, April 12–14, 2023 Proceedings. editor / João Correia ; Stephen Smith ; Raneem Qaddoura. Gewerbestrasse : Springer, 2023. pp. 656-671 (Lecture Notes in Computer Science).

@inproceedings{faa8ac32fc0d420689c1f8c9e67e19f9,

title = "Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers",

abstract = "Feature selection is becoming an essential part of machine learning pipelines, including the ones generated by recent AutoML tools. In case of datasets with epistatic interactions between the features, like many datasets from the bioinformatics domain, feature selection may even become crucial. A recent method called SLUG has outperformed the state-of-the-art algorithms for feature selection on a large set of epistatic noisy datasets. SLUG uses genetic programming (GP) as a classifier (learner), nested inside a genetic algorithm (GA) that performs feature selection (wrapper). In this work, we pair GA with different learners, in an attempt to match the results of SLUG with less computational effort. We also propose a new feedback mechanism between the learner and the wrapper to improve the convergence towards the key features. Although we do not match the results of SLUG, we demonstrate the positive effect of the feedback mechanism, motivating additional research in this area to further improve SLUG and other existing feature selection methods.",

keywords = "Feature Selection, Epistasis, Genetic Algorithms, Genetic Programming, Decision Trees, Machine Learning, Genome-Wide Association Studies",

author = "Pedro Carvalho and Bruno Ribeiro and Rodrigues, {Nuno M.} and Batista, {Jo{\~a}o E.} and Leonardo Vanneschi and Sara Silva",

note = "info:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UIDB%2F04152%2F2020/PT# Carvalho, P., Ribeiro, B., Rodrigues, N. M., Batista, J. E., Vanneschi, L., & Silva, S. (2023). Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers. In J. Correia, S. Smith, & R. Qaddoura (Eds.), Applications of Evolutionary Computation: 26th European Conference, EvoApplications 2023 Held as Part of EvoStar 2023 Brno, Czech Republic, April 12–14, 2023 Proceedings (pp. 656-671). (Lecture Notes in Computer Science; Vol. 13989). Springer. https://doi.org/10.1007/978-3-031-30229-9_42---This work was supported by FCT, Portugal, through funding of LASIGE Research Unit (UIDB/00408/2020, UIDP/00408/2020) and CISUC (UID/CEC/00326/2020); projects AICE (DSAIPA/DS/0113/2019), from FCT, and RETINA (NORTE-01-0145-FEDER-000062), supported by Norte Portugal Regional Operational Programme (NORTE 2020), under the PORTUGAL 2020 Partnership Agreement, through the European Regional Development Fund (ERDF). The authors acknowledge the work facilities and equipment provided by GECAD research center (UIDB/00760/2020) to the project team. The authors were also supported by their respective PhD grants, Pedro Carvalho (UI/BD/151053/2021), Nuno Rodrigues (2021/05322/BD), Jo{\~a}o Batista (SFRH/BD/143972/2019).; 26th International Conference on the Applications of Evolutionary Computation, EvoApplications 2023 ; Conference date: 12-04-2023 Through 14-04-2023",

year = "2023",

month = apr,

doi = "10.1007/978-3-031-30229-9_42",

language = "English",

isbn = "978-3-031-30228-2",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "656--671",

editor = "Jo{\~a}o Correia and Stephen Smith and Raneem Qaddoura",

booktitle = "Applications of Evolutionary Computation",

address = "Netherlands",

url = "https://www.evostar.org/2023/evoapps/",

}

Carvalho, P, Ribeiro, B, Rodrigues, NM, Batista, JE, Vanneschi, L & Silva, S 2023, Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers. in J Correia, S Smith & R Qaddoura (eds), Applications of Evolutionary Computation: 26th European Conference, EvoApplications 2023 Held as Part of EvoStar 2023 Brno, Czech Republic, April 12–14, 2023 Proceedings. Lecture Notes in Computer Science, vol. 13989, Springer, Gewerbestrasse, pp. 656-671, 26th International Conference on the Applications of Evolutionary Computation, Brno, Czech Republic, 12/04/23. https://doi.org/10.1007/978-3-031-30229-9_42

Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers. / Carvalho, Pedro; Ribeiro, Bruno; Rodrigues, Nuno M. et al.
Applications of Evolutionary Computation: 26th European Conference, EvoApplications 2023 Held as Part of EvoStar 2023 Brno, Czech Republic, April 12–14, 2023 Proceedings. ed. / João Correia; Stephen Smith; Raneem Qaddoura. Gewerbestrasse: Springer, 2023. p. 656-671 (Lecture Notes in Computer Science; Vol. 13989).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers

AU - Carvalho, Pedro

AU - Ribeiro, Bruno

AU - Rodrigues, Nuno M.

AU - Batista, João E.

AU - Vanneschi, Leonardo

AU - Silva, Sara

N1 - Conference code: 26

PY - 2023/4

Y1 - 2023/4

N2 - Feature selection is becoming an essential part of machine learning pipelines, including the ones generated by recent AutoML tools. In case of datasets with epistatic interactions between the features, like many datasets from the bioinformatics domain, feature selection may even become crucial. A recent method called SLUG has outperformed the state-of-the-art algorithms for feature selection on a large set of epistatic noisy datasets. SLUG uses genetic programming (GP) as a classifier (learner), nested inside a genetic algorithm (GA) that performs feature selection (wrapper). In this work, we pair GA with different learners, in an attempt to match the results of SLUG with less computational effort. We also propose a new feedback mechanism between the learner and the wrapper to improve the convergence towards the key features. Although we do not match the results of SLUG, we demonstrate the positive effect of the feedback mechanism, motivating additional research in this area to further improve SLUG and other existing feature selection methods.

AB - Feature selection is becoming an essential part of machine learning pipelines, including the ones generated by recent AutoML tools. In case of datasets with epistatic interactions between the features, like many datasets from the bioinformatics domain, feature selection may even become crucial. A recent method called SLUG has outperformed the state-of-the-art algorithms for feature selection on a large set of epistatic noisy datasets. SLUG uses genetic programming (GP) as a classifier (learner), nested inside a genetic algorithm (GA) that performs feature selection (wrapper). In this work, we pair GA with different learners, in an attempt to match the results of SLUG with less computational effort. We also propose a new feedback mechanism between the learner and the wrapper to improve the convergence towards the key features. Although we do not match the results of SLUG, we demonstrate the positive effect of the feedback mechanism, motivating additional research in this area to further improve SLUG and other existing feature selection methods.

KW - Feature Selection

KW - Epistasis

KW - Genetic Algorithms

KW - Genetic Programming

KW - Decision Trees

KW - Machine Learning

KW - Genome-Wide Association Studies

UR - http://www.scopus.com/inward/record.url?scp=85159438362&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-30229-9_42

DO - 10.1007/978-3-031-30229-9_42

M3 - Conference contribution

SN - 978-3-031-30228-2

T3 - Lecture Notes in Computer Science

SP - 656

EP - 671

BT - Applications of Evolutionary Computation

A2 - Correia, João

A2 - Smith, Stephen

A2 - Qaddoura, Raneem

PB - Springer

CY - Gewerbestrasse

T2 - 26th International Conference on the Applications of Evolutionary Computation

Y2 - 12 April 2023 through 14 April 2023

ER -

Carvalho P, Ribeiro B, Rodrigues NM, Batista JE, Vanneschi L, Silva S. Feature Selection on Epistatic Problems Using Genetic Algorithms with Nested Classifiers. In Correia J, Smith S, Qaddoura R, editors, Applications of Evolutionary Computation: 26th European Conference, EvoApplications 2023 Held as Part of EvoStar 2023 Brno, Czech Republic, April 12–14, 2023 Proceedings. Gewerbestrasse: Springer. 2023. p. 656-671. (Lecture Notes in Computer Science). doi: 10.1007/978-3-031-30229-9_42