scholarly journals Hybrid feature selection method based on particle swarm optimization and adaptive local search method

2021 ◽  
Vol 11 (3) ◽  
pp. 2414
Author(s):  
Malek Alzaqebah ◽  
Sana Jawarneh ◽  
Rami Mustafa A. Mohammad ◽  
Mutasem K. Alsmadi ◽  
Ibrahim Al-marashdeh ◽  
...  
Keyword(s):  
Feature Selection ◽  
Particle Swarm Optimization ◽  
Local Search ◽  
Particle Swarm ◽  
Feature Selection Method ◽  
Pso Algorithm ◽  
Search Space ◽  
Swarm Optimization ◽  
Candidate Solution ◽  
Using Data

Machine learning has been expansively examined with data classification as the most popularly researched subject. The accurateness of prediction is impacted by the data provided to the classification algorithm. Meanwhile, utilizing a large amount of data may incur costs especially in data collection and preprocessing. Studies on feature selection were mainly to establish techniques that can decrease the number of utilized features (attributes) in classification, also using data that generate accurate prediction is important. Hence, a particle swarm optimization (PSO) algorithm is suggested in the current article for selecting the ideal set of features. PSO algorithm showed to be superior in different domains in exploring the search space and local search algorithms are good in exploiting the search regions. Thus, we propose the hybridized PSO algorithm with an adaptive local search technique which works based on the current PSO search state and used for accepting the candidate solution. Having this combination balances the local intensification as well as the global diversification of the searching process. Hence, the suggested algorithm surpasses the original PSO algorithm and other comparable approaches, in terms of performance.

scholarly journals On the Performance of Linear Decreasing Inertia Weight Particle Swarm Optimization for Global Optimization

2013 ◽  
Vol 2013 ◽  
pp. 1-12 ◽  
Author(s):  
Martins Akugbe Arasomwan ◽  
Aderemi Oluyinka Adewumi
Keyword(s):  
Global Optimization ◽  
Particle Swarm Optimization ◽  
Optimization Problems ◽  
Particle Swarm ◽  
Pso Algorithm ◽  
Search Space ◽  
Terminal Point ◽  
Simulation Experiments ◽  
Swarm Optimization ◽  
Inertia Weight

Linear decreasing inertia weight (LDIW) strategy was introduced to improve on the performance of the original particle swarm optimization (PSO). However, linear decreasing inertia weight PSO (LDIW-PSO) algorithm is known to have the shortcoming of premature convergence in solving complex (multipeak) optimization problems due to lack of enough momentum for particles to do exploitation as the algorithm approaches its terminal point. Researchers have tried to address this shortcoming by modifying LDIW-PSO or proposing new PSO variants. Some of these variants have been claimed to outperform LDIW-PSO. The major goal of this paper is to experimentally establish the fact that LDIW-PSO is very much efficient if its parameters are properly set. First, an experiment was conducted to acquire a percentage value of the search space limits to compute the particle velocity limits in LDIW-PSO based on commonly used benchmark global optimization problems. Second, using the experimentally obtained values, five well-known benchmark optimization problems were used to show the outstanding performance of LDIW-PSO over some of its competitors which have in the past claimed superiority over it. Two other recent PSO variants with different inertia weight strategies were also compared with LDIW-PSO with the latter outperforming both in the simulation experiments conducted.

An Improved Particle Swarm Optimization Algorithm for Global Multidimensional Optimization

2017 ◽  
Vol 29 (1) ◽  
pp. 127-142
Author(s):  
Rkia Fajr ◽  
Abdelaziz Bouroumi
Keyword(s):  
Particle Swarm Optimization ◽  
Learning Strategies ◽  
Particle Swarm ◽  
Pso Algorithm ◽  
Search Space ◽  
Human Learning ◽  
Computing Methods ◽  
Swarm Optimization ◽  
Starting Point ◽  
New Variant

Abstract This paper introduces a new variant of the particle swarm optimization (PSO) algorithm, designed for global optimization of multidimensional functions. The goal of this variant, called ImPSO, is to improve the exploration and exploitation abilities of the algorithm by introducing a new operation in the iterative search process. The use of this operation is governed by a stochastic rule that ensures either the exploration of new regions of the search space or the exploitation of good intermediate solutions. The proposed method is inspired by collaborative human learning and uses as a starting point a basic PSO variant with constriction factor and velocity clamping. Simulation results that show the ability of ImPSO to locate the global optima of multidimensional functions are presented for 10 well-know benchmark functions from CEC-2013 and CEC-2005. These results are compared with the PSO variant used as starting point, three other PSO variants, one of which is based on human learning strategies, and three alternative evolutionary computing methods.

scholarly journals Particle Swarm Optimization

2017 ◽  
pp. 614-635 ◽  
Author(s):  
Alaa Tharwat ◽  
Tarek Gaber ◽  
Aboul Ella Hassanien ◽  
Basem E. Elnaghi
Keyword(s):  
Particle Swarm Optimization ◽  
Optimization Algorithms ◽  
Particle Swarm ◽  
Optimal Solution ◽  
Parameter Tuning ◽  
Pso Algorithm ◽  
Search Space ◽  
Learning Problems ◽  
Swarm Optimization ◽  
Numerical Example

Optimization algorithms are necessary to solve many problems such as parameter tuning. Particle Swarm Optimization (PSO) is one of these optimization algorithms. The aim of PSO is to search for the optimal solution in the search space. This paper highlights the basic background needed to understand and implement the PSO algorithm. This paper starts with basic definitions of the PSO algorithm and how the particles are moved in the search space to find the optimal or near optimal solution. Moreover, a numerical example is illustrated to show how the particles are moved in a convex optimization problem. Another numerical example is illustrated to show how the PSO trapped in a local minima problem. Two experiments are conducted to show how the PSO searches for the optimal parameters in one-dimensional and two-dimensional spaces to solve machine learning problems.

scholarly journals Variable-Length Particle Swarm Optimization for Feature Selection on High-Dimensional Classification

2021 ◽  
Author(s):  
B Tran ◽  
Bing Xue ◽  
Mengjie Zhang
Keyword(s):  
Feature Selection ◽  
Particle Swarm Optimization ◽  
Computational Cost ◽  
Particle Swarm ◽  
Search Space ◽  
Classification Performance ◽  
Variable Length ◽  
High Dimensional ◽  
Swarm Optimization ◽  
Local Optima

© 1997-2012 IEEE. With a global search mechanism, particle swarm optimization (PSO) has shown promise in feature selection (FS). However, most of the current PSO-based FS methods use a fix-length representation, which is inflexible and limits the performance of PSO for FS. When applying these methods to high-dimensional data, it not only consumes a significant amount of memory but also requires a high computational cost. Overcoming this limitation enables PSO to work on data with much higher dimensionality which has become more and more popular with the advance of data collection technologies. In this paper, we propose the first variable-length PSO representation for FS, enabling particles to have different and shorter lengths, which defines smaller search space and therefore, improves the performance of PSO. By rearranging features in a descending order of their relevance, we facilitate particles with shorter lengths to achieve better classification performance. Furthermore, using the proposed length changing mechanism, PSO can jump out of local optima, further narrow the search space and focus its search on smaller and more fruitful area. These strategies enable PSO to reach better solutions in a shorter time. Results on ten high-dimensional datasets with varying difficulties show that the proposed variable-length PSO can achieve much smaller feature subsets with significantly higher classification performance in much shorter time than the fixed-length PSO methods. The proposed method also outperformed the compared non-PSO FS methods in most cases. © 2019 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Sign in / Sign up

Export Citation Format

Share Document