Binary Restructuring Particle Swarm Optimization and Its Application

Abstract

Restructuring Particle Swarm Optimization (RPSO) algorithm has been developed as an intelligent approach based on the linear system theory of particle swarm optimization (PSO). It streamlines the flow of the PSO algorithm, specifically targeting continuous optimization problems. In order to adapt RPSO for solving discrete optimization problems, this paper proposes the binary Restructuring Particle Swarm Optimization (BRPSO) algorithm. Unlike other binary metaheuristic algorithms, BRPSO does not utilize the transfer function. The particle updating process in BRPSO relies solely on comparison results between values derived from the position updating formula and a random number. Additionally, a novel perturbation term is incorporated into the position updating formula of BRPSO. Notably, BRPSO requires fewer parameters and exhibits high exploration capability during the early stages. To evaluate the efficacy of BRPSO, comprehensive experiments are conducted by comparing it against four peer algorithms in the context of feature selection problems. The experimental results highlight the competitive nature of BRPSO in terms of both classification accuracy and the number of selected features.

Automated Collective Variable Discovery for MFSD2A transporter from molecular dynamics simulations

“…The fitness of each particle, determined by the classification accuracy and the size of a feature subset, is then evaluated. In this study, the following fitness function 55 is used: where P is the classification accuracy, N sel is the number of selected features, N feat is the total number of features available, α ∈ [0,1] is a weighting factor that balances classifier performance and subset size ( α was set to 0.99 based on the previous work) 57-59 . After evaluating the fitness of the entire swarm, each particle’s velocity is updated according to the following equation: where t is the iteration index, X ( t ) is the particle position, p best is the particle’s historical best position, g best is the historical best position of the entire swarm, r 1 ( t ) and r 2 ( t ) are two random vectors in [0, 1] range that share the same dimensionality with the feature space.…”

“…where 𝑃 is the classification accuracy, 𝑁 is the number of selected features, 𝑁 is the total number of features available, 𝛼 ∈ 0,1 is a weighting factor that balances classifier performance and subset size (𝛼 was set to 0.99 based on the previous work) [57][58][59] . After evaluating the fitness of the entire swarm, each particle's velocity is updated according to the following equation:…”

An Advanced Bio-Inspired Mantis Search Algorithm for Characterization of PV Panel and Global Optimization of Its Model Parameters

“…However, there is no guarantee that the methods utilised will yield such an ideal result. Thus, the solution generated by an optimiser for any particular issue is known as a quasi-optimal [31]. Metaheuristic strategies require being capable of executing and overseeing searches at the global, as well as local levels, in order to organise a successful investigation in the problem-solving domain.…”

A Subtraction-Average-Based Optimizer for Solving Engineering Problems with Applications on TCSC Allocation in Power Systems

“…Unfortunately, there is no assurance, however, whether the algorithms being used will exactly produce such an optimal solution. As a result, the solution produced by an optimization method for a problem is referred to as a quasi-optimal, which may or may not be equivalent to the global optimum [ 6 ]. To organize an effective search in the problem-solving area, metaheuristic techniques must be capable of executing and overseeing queries at the local as well as global levels.…”