Feature Selection by Hybrid Binary Ant Lion Optimizer with COVID-19 dataset
Ivana Strumberger, Andjela Rakic, Stefan Stanojlovic, Jelena Arandjelovic, Timea Bezdan, Miodrag Živković, Nebojša Bačanin
Abstract
Metaheuristic optimization is becoming increasingly popular as a way for addressing complicated issues that are difficult to tackle using standard approaches. The capacity to solve issues that may occur as a result of the interactions of basic information processing units is known as swarm intelligence. Swarm intelligence is a type of artificial intelligence methods, that is based on indirect communications between individual from population. The ant lion optimizer algorithm is recently proposed swarm intelligence algorithm. However, basic ant lion version suffers from some deficiencies which are addressed in this study by performing hybridization with well-known firefly algorithm. Proposed hybrid approach was adapted for solving important feature selection problem from machine learning domain. Hybrid metaheuristics was first validated against 10 UCL datasets and afterwards it was applied to a novel COVID-19 dataset. Moreover, comparative analysis with similar methods tested under the same condition was presented. Obtained experimental results prove the efficiency of proposed hybrid ant lion optimizer for tackling feature selection challenge.