Cargando…

Binary Bamboo Forest Growth Optimization Algorithm for Feature Selection Problem

Inspired by the bamboo growth process, Chu et al. proposed the Bamboo Forest Growth Optimization (BFGO) algorithm. It incorporates bamboo whip extension and bamboo shoot growth into the optimization process. It can be applied very well to classical engineering problems. However, binary values can on...

Descripción completa

Detalles Bibliográficos
Autores principales: Pan, Jeng-Shyang, Yue, Longkang, Chu, Shu-Chuan, Hu, Pei, Yan, Bin, Yang, Hongmei
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9955014/
https://www.ncbi.nlm.nih.gov/pubmed/36832680
http://dx.doi.org/10.3390/e25020314
_version_ 1784894252326584320
author Pan, Jeng-Shyang
Yue, Longkang
Chu, Shu-Chuan
Hu, Pei
Yan, Bin
Yang, Hongmei
author_facet Pan, Jeng-Shyang
Yue, Longkang
Chu, Shu-Chuan
Hu, Pei
Yan, Bin
Yang, Hongmei
author_sort Pan, Jeng-Shyang
collection PubMed
description Inspired by the bamboo growth process, Chu et al. proposed the Bamboo Forest Growth Optimization (BFGO) algorithm. It incorporates bamboo whip extension and bamboo shoot growth into the optimization process. It can be applied very well to classical engineering problems. However, binary values can only take 0 or 1, and for some binary optimization problems, the standard BFGO is not applicable. This paper firstly proposes a binary version of BFGO, called BBFGO. By analyzing the search space of BFGO under binary conditions, the new curve V-shaped and Taper-shaped transfer function for converting continuous values into binary BFGO is proposed for the first time. A long-mutation strategy with a new mutation approach is presented to solve the algorithmic stagnation problem. Binary BFGO and the long-mutation strategy with a new mutation are tested on 23 benchmark test functions. The experimental results show that binary BFGO achieves better results in solving the optimal values and convergence speed, and the variation strategy can significantly enhance the algorithm’s performance. In terms of application, 12 data sets derived from the UCI machine learning repository are selected for feature-selection implementation and compared with the transfer functions used by BGWO-a, BPSO-TVMS and BQUATRE, which demonstrates binary BFGO algorithm’s potential to explore the attribute space and choose the most significant features for classification issues.
format Online
Article
Text
id pubmed-9955014
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-99550142023-02-25 Binary Bamboo Forest Growth Optimization Algorithm for Feature Selection Problem Pan, Jeng-Shyang Yue, Longkang Chu, Shu-Chuan Hu, Pei Yan, Bin Yang, Hongmei Entropy (Basel) Article Inspired by the bamboo growth process, Chu et al. proposed the Bamboo Forest Growth Optimization (BFGO) algorithm. It incorporates bamboo whip extension and bamboo shoot growth into the optimization process. It can be applied very well to classical engineering problems. However, binary values can only take 0 or 1, and for some binary optimization problems, the standard BFGO is not applicable. This paper firstly proposes a binary version of BFGO, called BBFGO. By analyzing the search space of BFGO under binary conditions, the new curve V-shaped and Taper-shaped transfer function for converting continuous values into binary BFGO is proposed for the first time. A long-mutation strategy with a new mutation approach is presented to solve the algorithmic stagnation problem. Binary BFGO and the long-mutation strategy with a new mutation are tested on 23 benchmark test functions. The experimental results show that binary BFGO achieves better results in solving the optimal values and convergence speed, and the variation strategy can significantly enhance the algorithm’s performance. In terms of application, 12 data sets derived from the UCI machine learning repository are selected for feature-selection implementation and compared with the transfer functions used by BGWO-a, BPSO-TVMS and BQUATRE, which demonstrates binary BFGO algorithm’s potential to explore the attribute space and choose the most significant features for classification issues. MDPI 2023-02-08 /pmc/articles/PMC9955014/ /pubmed/36832680 http://dx.doi.org/10.3390/e25020314 Text en © 2023 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Pan, Jeng-Shyang
Yue, Longkang
Chu, Shu-Chuan
Hu, Pei
Yan, Bin
Yang, Hongmei
Binary Bamboo Forest Growth Optimization Algorithm for Feature Selection Problem
title Binary Bamboo Forest Growth Optimization Algorithm for Feature Selection Problem
title_full Binary Bamboo Forest Growth Optimization Algorithm for Feature Selection Problem
title_fullStr Binary Bamboo Forest Growth Optimization Algorithm for Feature Selection Problem
title_full_unstemmed Binary Bamboo Forest Growth Optimization Algorithm for Feature Selection Problem
title_short Binary Bamboo Forest Growth Optimization Algorithm for Feature Selection Problem
title_sort binary bamboo forest growth optimization algorithm for feature selection problem
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9955014/
https://www.ncbi.nlm.nih.gov/pubmed/36832680
http://dx.doi.org/10.3390/e25020314
work_keys_str_mv AT panjengshyang binarybambooforestgrowthoptimizationalgorithmforfeatureselectionproblem
AT yuelongkang binarybambooforestgrowthoptimizationalgorithmforfeatureselectionproblem
AT chushuchuan binarybambooforestgrowthoptimizationalgorithmforfeatureselectionproblem
AT hupei binarybambooforestgrowthoptimizationalgorithmforfeatureselectionproblem
AT yanbin binarybambooforestgrowthoptimizationalgorithmforfeatureselectionproblem
AT yanghongmei binarybambooforestgrowthoptimizationalgorithmforfeatureselectionproblem