Open AccessA Proposed Intelligent Features Selection Method Using Meerkat Clan Algorithm
Author(s) -
Noor Ghazi M. Jameel,
Hasanen S. Abdullah
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1804/1/012061
Subject(s) - feature selection , computer science , artificial intelligence , feature (linguistics) , ant colony optimization algorithms , pattern recognition (psychology) , selection (genetic algorithm) , swarm behaviour , data mining , set (abstract data type) , fitness function , fuzzy logic , ant colony , machine learning , genetic algorithm , philosophy , linguistics , programming language
The feature selection process stand to select the subset of feature from the set of datasets to get the relevant an important features and remove irrelevant features. The feature selection types are supervised, unsupervised or semi-supervised. The feature selection methods are divided into: traditional methods such as Correlation and Information Gain or others and intelligent methods such as fuzzy logic, or use the swarm intelligence methods in feature selection such as Ant colony and Bees colony. This paper proposes an Intelligent Features Selection method based on Meerkat Clan Algorithm (IFS\MCA) system, when it used the modern MCA as first time to use the algorithm in feature selection to select the best subset of feature. This proposal used four set of dataset which as (Fri_c4, Sonar, Scene, Satellite), the efficiency of the proposal in feature selection is powerful and high accuracy when comparing with other standard methods in feature selection which as Correlation and Information gain when use the same set of dataset. The proposed IFS\MCA system uses the Mean Absolute Deviation (MAD) as a proposed choice as a fitness function (which is used as a first time with feature selection process).