Improved Slime Mould Algorithm based on Firefly Algorithm for feature selection: A case study on QSAR model

Feature selection (FS) methods are necessary to develop intelligent analysis tools that require data preprocessing and enhancing the performance of the machine learning algorithms. FS aims to maximize the classification accuracy by minimizing the number of selected features. This paper presents a ne...

Full description

Saved in:
Bibliographic Details
Published in:Engineering with computers 2022-08, Vol.38 (Suppl 3), p.2407-2421
Main Authors: Ewees, Ahmed A., Abualigah, Laith, Yousri, Dalia, Algamal, Zakariya Yahya, Al-qaness, Mohammed A. A., Ibrahim, Rehab Ali, Abd Elaziz, Mohamed
Format: Article
Language:English
Subjects:
Algorithms
CAE) and Design
Calculus of Variations and Optimal Control
Optimization
Case studies
Classical Mechanics
Classification
Computer engineering
Computer Science
Computer-Aided Engineering (CAD
Computers
Control
Datasets
Feature selection
Genetic algorithms
Heuristic methods
Informatics
Internet of Things
Machine learning
Math. Applications in Chemistry
Mathematical and Computational Engineering
Medical research
Methods
Optimization algorithms
Original Article
Slime
Systems Theory
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Feature selection (FS) methods are necessary to develop intelligent analysis tools that require data preprocessing and enhancing the performance of the machine learning algorithms. FS aims to maximize the classification accuracy by minimizing the number of selected features. This paper presents a new FS method using a modified Slime mould algorithm (SMA) based on the firefly algorithm (FA). In the developed SMAFA, FA is adopted to improve the exploration of SMA, since it has high ability to discover the feasible regions which have optima solution. This will lead to enhance the convergence by increasing the quality of the final output. SMAFA is evaluated using twenty UCI datasets and also with comprehensive comparisons to a number of the existing MH algorithms. To further assess the applicability of SMAFA, two high-dimensional datasets related to the QSAR modeling are used. Experimental results verified the promising performance of SMAFA using different performance measures.
ISSN:0177-0667
1435-5663
DOI:10.1007/s00366-021-01342-6