Global chaotic bat algorithm for feature selection

The wrapper algorithm adopts the performance of the learning algorithm as the evaluation criteria to obtain excellent classification performance. However, the wrapper algorithm is prone to converge prematurely. A global chaotic bat algorithm (GCBA) is put up forward to improve this shortage. First,...

Full description

Saved in:
Bibliographic Details
Published in:The Journal of supercomputing Vol. 78; no. 17; pp. 18754 - 18776
Main Authors: Li, Ying, Cui, Xueting, Fan, Jiahao, Wang, Tan
Format: Journal Article
Language:English
Published: New York Springer US 01.11.2022
Springer Nature B.V
Subjects:
Adaptive learning
Algorithms
Classification
Compilers
Computer Science
Datasets
Gene expression
Interpreters
Machine learning
Processor Architectures
Programming Languages
Solution space
Transfer functions
Wrapper feature selection algorithm
Feature selection
Bat algorithm
Chaotic map
Classification
ISSN:0920-8542, 1573-0484
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The wrapper algorithm adopts the performance of the learning algorithm as the evaluation criteria to obtain excellent classification performance. However, the wrapper algorithm is prone to converge prematurely. A global chaotic bat algorithm (GCBA) is put up forward to improve this shortage. First, GCBA applies chaotic map to population initialization to cover the entire solution space. In addition, adaptive learning factors are presented to balance exploration and exploration. The learning factor of local optimal position gradually decreases in the early stage while the learning factor of global optimal position gradually increases in the later stage. Finally, to improve the exploitation, an improved transfer function is proposed, which transfers the continuous space to discrete binary space. GCBA is tested on 14 UCI data sets and 5 gene expression data sets compared with other 6 comparison algorithms. Compared with other algorithms, the results show that GCBA is able to achieve better classification performance.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0920-8542
1573-0484
DOI:10.1007/s11227-022-04606-0