Feature Selection Using Enhanced Particle Swarm Optimisation for Classification Models

In this research, we propose two Particle Swarm Optimisation (PSO) variants to undertake feature selection tasks. The aim is overcome major shortcomings of the original PSO model, i.e., premature convergence and weak exploitation around near optimal solutions. first proposed variant incorporates four key operations, including a modified operation with rectified personal global best signals, spiral search based local exploitation, Gaussian distribution-based swarm leader enhancement, mirroring mutation operations for worst solution improvement. second model enhances one through new strategies, an adaptive exemplar breeding mechanism incorporating multiple nonlinear function oriented coefficients, exponential scattering schemes leader, respectively. comparison set 15 classical advanced methods, models illustrate statistical superiority discriminative total 13 data sets.