Chapter 1
A New Hybrid Binary Algorithm of Bat
Algorithm and Differential Evolution for
Feature Selection and Classification
Abdelmonem M. Ibrahim and Mohamed A. Tawhid
1 Introduction
Feature selection (FS) is an approach for figuring out the most essential features
and eliminating irrelevant and redundant data [1, 2]. The goals of FS are to reduce
the dimensionality of the data, enhance the prediction accuracy, and perceive data
for various applications in machine learning [3]. In a variety of applications, data
representation often employs many features with some redundant ones, which means
selected essential features can be considered while the irrelevant features (superflu-
ous) can be eliminated. Also, the output is affected by the pertinent features because
they provide useful information concerning the data, and the outcomes will be unclear
if any of them is kept out [4]. The standard optimization methods have some draw-
backs in solving the feature selection problems because for N features, 2
N
feature
subsets have to be generated and computed for a dataset, and it is known to be an
NP-hard problem [5]. Thus, metaheuristic algorithms (MAs) are the alternative for
overcoming these drawbacks and searching for the optimum solution [4, 6]. MAs
may be inspired by biological interaction, group dynamics, nature, and social behav-
ior. The binary version of these algorithms allows many researchers to deal with
complex problems like feature selection and attain good results. MAs have been uti-
lized to solve the feature selection problem, for example, binary bat algorithm (BBA)
[7], binary crow search algorithm (CSA) [8], binary gray wolf optimization (bGWO)
A. M. Ibrahim
Department of Mathematics, Faculty of Science, Al-Azhar University,
Assiut Branch, Asyut, Egypt
e-mail: abdelmonem@azhar.edu.eg
M. A. Tawhid (B)
Department of Mathematics and Statistics, Faculty of Science,
Thompson Rivers University, Kamloops, BC V2C 0C8, Canada
e-mail: Mtawhid@tru.ca
© The Editor(s) (if applicable) and The Author(s), under exclusive license
to Springer Nature Singapore Pte Ltd. 2021
N. Dey and V. Rajinikanth (eds.), Applications of Bat Algorithm and its Variants,
Springer Tracts in Nature-Inspired Computing,
https://doi.org/10.1007/978-981-15-5097-3_1
1