Optimizing Feature Selection for Binary Classification with Noisy Labels : A Genetic Algorithm Approach

Vandad Imani; Elaheh Moradi; Carlos Sevilla-Salcedo; Vittorio Fortino; Jussi Tohka

doi:10.1007/978-3-031-56950-0_33

Optimizing Feature Selection for Binary Classification with Noisy Labels : A Genetic Algorithm Approach

Vandad Imani^*, Elaheh Moradi, Carlos Sevilla-Salcedo, Vittorio Fortino, Jussi Tohka

^*Corresponding author for this work

Department of Computer Science

University of Eastern Finland

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

Abstract

Feature selection in noisy label scenarios remains an understudied topic. We propose a novel genetic algorithm-based approach, the Noise-Aware Multi-Objective Feature Selection Genetic Algorithm (NMFS-GA), for selecting optimal feature subsets in binary classification with noisy labels. NMFS-GA offers a unified framework for selecting feature subsets that are both accurate and interpretable. We evaluate NMFS-GA on synthetic datasets with label noise, a Breast Cancer dataset enriched with noisy features, and a real-world ADNI dataset for dementia conversion prediction. Our results indicate that NMFS-GA can effectively select feature subsets that improve the accuracy and interpretability of binary classifiers in scenarios with noisy labels.

Original language	English
Title of host publication	Proceedings of the 2nd International Conference on Advances in Computing Research, ACR 2024
Editors	Kevin Daimi, Abeer Al Sadoon
Publisher	Springer
Pages	392-403
Number of pages	12
ISBN (Electronic)	978-3-031-56950-0
ISBN (Print)	978-3-031-56949-4
DOIs	https://doi.org/10.1007/978-3-031-56950-0_33
Publication status	Published - 29 Mar 2024
MoE publication type	A4 Conference publication
Event	International Conference on Advances in Computing Research - Madrid, Spain Duration: 3 Jun 2024 → 5 Jun 2024 Conference number: 2

Publication series

Name	Lecture Notes in Networks and Systems
Publisher	Springer
Volume	956 LNNS
ISSN (Print)	2367-3370
ISSN (Electronic)	2367-3389

Conference

Conference	International Conference on Advances in Computing Research
Abbreviated title	ACR
Country/Territory	Spain
City	Madrid
Period	03/06/2024 → 05/06/2024

Keywords

ADNI
Alzheimer’s disease
Classification
Converter dementia
Feature selection
Genetic algorithm
Magnetic resonance imaging
Mild cognitive impairment
Noisy labels

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1007/978-3-031-56950-0_33

https://arxiv.org/abs/2401.06546

Cite this

Imani, V., Moradi, E., Sevilla-Salcedo, C., Fortino, V., & Tohka, J. (2024). Optimizing Feature Selection for Binary Classification with Noisy Labels : A Genetic Algorithm Approach. In K. Daimi, & A. Al Sadoon (Eds.), Proceedings of the 2nd International Conference on Advances in Computing Research, ACR 2024 (pp. 392-403). (Lecture Notes in Networks and Systems; Vol. 956 LNNS). Springer. https://doi.org/10.1007/978-3-031-56950-0_33

Imani, Vandad ; Moradi, Elaheh ; Sevilla-Salcedo, Carlos et al. / Optimizing Feature Selection for Binary Classification with Noisy Labels : A Genetic Algorithm Approach. Proceedings of the 2nd International Conference on Advances in Computing Research, ACR 2024. editor / Kevin Daimi ; Abeer Al Sadoon. Springer, 2024. pp. 392-403 (Lecture Notes in Networks and Systems).

@inproceedings{d394bfea1c2b4efdbbb589fe0fc6d7a0,

title = "Optimizing Feature Selection for Binary Classification with Noisy Labels : A Genetic Algorithm Approach",

abstract = "Feature selection in noisy label scenarios remains an understudied topic. We propose a novel genetic algorithm-based approach, the Noise-Aware Multi-Objective Feature Selection Genetic Algorithm (NMFS-GA), for selecting optimal feature subsets in binary classification with noisy labels. NMFS-GA offers a unified framework for selecting feature subsets that are both accurate and interpretable. We evaluate NMFS-GA on synthetic datasets with label noise, a Breast Cancer dataset enriched with noisy features, and a real-world ADNI dataset for dementia conversion prediction. Our results indicate that NMFS-GA can effectively select feature subsets that improve the accuracy and interpretability of binary classifiers in scenarios with noisy labels.",

keywords = "ADNI, Alzheimer{\textquoteright}s disease, Classification, Converter dementia, Feature selection, Genetic algorithm, Magnetic resonance imaging, Mild cognitive impairment, Noisy labels",

author = "Vandad Imani and Elaheh Moradi and Carlos Sevilla-Salcedo and Vittorio Fortino and Jussi Tohka",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.; International Conference on Advances in Computing Research, ACR ; Conference date: 03-06-2024 Through 05-06-2024",

year = "2024",

month = mar,

day = "29",

doi = "10.1007/978-3-031-56950-0_33",

language = "English",

isbn = "978-3-031-56949-4",

series = "Lecture Notes in Networks and Systems",

publisher = "Springer",

pages = "392--403",

editor = "Kevin Daimi and {Al Sadoon}, Abeer",

booktitle = "Proceedings of the 2nd International Conference on Advances in Computing Research, ACR 2024",

address = "Germany",

}

Imani, V, Moradi, E, Sevilla-Salcedo, C, Fortino, V & Tohka, J 2024, Optimizing Feature Selection for Binary Classification with Noisy Labels : A Genetic Algorithm Approach. in K Daimi & A Al Sadoon (eds), Proceedings of the 2nd International Conference on Advances in Computing Research, ACR 2024. Lecture Notes in Networks and Systems, vol. 956 LNNS, Springer, pp. 392-403, International Conference on Advances in Computing Research, Madrid, Spain, 03/06/2024. https://doi.org/10.1007/978-3-031-56950-0_33

Optimizing Feature Selection for Binary Classification with Noisy Labels : A Genetic Algorithm Approach. / Imani, Vandad; Moradi, Elaheh; Sevilla-Salcedo, Carlos et al.
Proceedings of the 2nd International Conference on Advances in Computing Research, ACR 2024. ed. / Kevin Daimi; Abeer Al Sadoon. Springer, 2024. p. 392-403 (Lecture Notes in Networks and Systems; Vol. 956 LNNS).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - Optimizing Feature Selection for Binary Classification with Noisy Labels : A Genetic Algorithm Approach

AU - Imani, Vandad

AU - Moradi, Elaheh

AU - Sevilla-Salcedo, Carlos

AU - Fortino, Vittorio

AU - Tohka, Jussi

N1 - Conference code: 2

PY - 2024/3/29

Y1 - 2024/3/29

N2 - Feature selection in noisy label scenarios remains an understudied topic. We propose a novel genetic algorithm-based approach, the Noise-Aware Multi-Objective Feature Selection Genetic Algorithm (NMFS-GA), for selecting optimal feature subsets in binary classification with noisy labels. NMFS-GA offers a unified framework for selecting feature subsets that are both accurate and interpretable. We evaluate NMFS-GA on synthetic datasets with label noise, a Breast Cancer dataset enriched with noisy features, and a real-world ADNI dataset for dementia conversion prediction. Our results indicate that NMFS-GA can effectively select feature subsets that improve the accuracy and interpretability of binary classifiers in scenarios with noisy labels.

AB - Feature selection in noisy label scenarios remains an understudied topic. We propose a novel genetic algorithm-based approach, the Noise-Aware Multi-Objective Feature Selection Genetic Algorithm (NMFS-GA), for selecting optimal feature subsets in binary classification with noisy labels. NMFS-GA offers a unified framework for selecting feature subsets that are both accurate and interpretable. We evaluate NMFS-GA on synthetic datasets with label noise, a Breast Cancer dataset enriched with noisy features, and a real-world ADNI dataset for dementia conversion prediction. Our results indicate that NMFS-GA can effectively select feature subsets that improve the accuracy and interpretability of binary classifiers in scenarios with noisy labels.

KW - ADNI

KW - Alzheimer’s disease

KW - Classification

KW - Converter dementia

KW - Feature selection

KW - Genetic algorithm

KW - Magnetic resonance imaging

KW - Mild cognitive impairment

KW - Noisy labels

UR - http://www.scopus.com/inward/record.url?scp=85190673174&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-56950-0_33

DO - 10.1007/978-3-031-56950-0_33

M3 - Conference article in proceedings

AN - SCOPUS:85190673174

SN - 978-3-031-56949-4

T3 - Lecture Notes in Networks and Systems

SP - 392

EP - 403

BT - Proceedings of the 2nd International Conference on Advances in Computing Research, ACR 2024

A2 - Daimi, Kevin

A2 - Al Sadoon, Abeer

PB - Springer

T2 - International Conference on Advances in Computing Research

Y2 - 3 June 2024 through 5 June 2024

ER -

Imani V, Moradi E, Sevilla-Salcedo C, Fortino V, Tohka J. Optimizing Feature Selection for Binary Classification with Noisy Labels : A Genetic Algorithm Approach. In Daimi K, Al Sadoon A, editors, Proceedings of the 2nd International Conference on Advances in Computing Research, ACR 2024. Springer. 2024. p. 392-403. (Lecture Notes in Networks and Systems). doi: 10.1007/978-3-031-56950-0_33