Particle Swarm Optimization (PSO)-based Feature Selection: An Approach for Improving Performance of Classification Model

Main Article Content

Engr. Dr. Oluyemisi Adenike Oyedemi
Prof. Iyabo Olukemi Awoyelu
Prof. Emmanuel Rotimi Adagunodo

Abstract

The Particle Swarm Optimization (PSO) algorithm was implemented in this paper for
feature selection in order to improve the performance of web phishing classification model. The
algorithm initialized a population of particles, each of which represented a possible feature subset,
and then iteratively investigated subsets of features. The particles are updated according to their
global and personal best locations in an effort to choose the best feature subset. A total of 36 features
were selected out of the downloaded dataset comprising of 48 features. The performance of the whole
dataset was measured against the performance of the selected dataset. It was observed that the performance
of the classifier improved across all the metrics. The accuracy, precision, recall and F1-score increased from
93% to 96%, 92% to 97%, 92% to 96% and 93% to 97% respectively. The results indicated that the PSO
algorithm effectively identified a subset of features that improved the performance of the classifier. The
PSO algorithm's optimal feature subset produces superior classification performance when
compared to using the whole dataset. Through feature selection, this method demonstrated how
well PSO works to improve classifier accuracy and efficiency across a range of classification tasks.
Keywords: Classifier Performance, Feature Selection, Machine Learning, Particle Swarm Optimization
(PSO), Optimization, Redundancy

Article Details

How to Cite
OYEDEMI, O., Awoyelu, I. O., & Adagunodo, E. R. (2025). Particle Swarm Optimization (PSO)-based Feature Selection: An Approach for Improving Performance of Classification Model . INFOCOMP Journal of Computer Science, 23(2). Retrieved from https://infocomp.dcc.ufla.br/index.php/infocomp/article/view/4945
Section
Machine Learning and Computational Intelligence

References

[1] Gheyas, et al. Feature subset selection in large dimensionality domains,” Pattern Recognition, 43(1): 5–13, 2010.

[2] Thabtah et.al. Prediction phase in associative classification mining, International Journal of Software Engineering and Knowledge Engineering, 21(6): 855–876, 2011.

[3] Waqas,et al. Feature subset selection using multi-objective genetic algorithms. In 13th International Conference on IEEE International Multi-Topic Conference (INMIC), 1–6, 2009.

[4] Ke, et al. A multiobjective ACO algorithm for rough feature selection. In: Second Pacific-Asia Conference on Circuits, Communications and System (PACCS). 1: 207–210, 2010.

[5] Dash, et al. Feature selection for classification. Intelligent Data Analysis. 1(4): 131– 156, 1997.

[6] Kennedy, et al. Matching algorithms to problems: an experimental test of the particle swarm and some genetic algorithms on the multimodal problem generator. In: IEEE World Congress on Computational Intelligence. 78–83., 1998.

[7] Moraglio, et al. Geometric Particle Swarm Optimisation. In: Genetic Programming. Springer, Lecture Note in Computer Science (LNCS), 4445: 125-136, 2007.

[8] Mohemmed, et al. Particle Swarm Optimization based Adaboost for face detection. In: IEEE Congress on Evolutionary Computation, 2494–2501, 2009.

[9] Wang, et al. A novel rough set reduction algorithm to feature selection based on artificial fish swarm algorithm, In Proceedings of the International Conference in Swarm Intelligence, Springer, 24-35, 2014.

[10] Hassan, et al. A comparison of particle swarm optimization and the genetic algorithm, In Proceedings of the 46thAIAA/ASME/ASCE/AHS/ASC Structures, Structural Dy-namics and Materials Conference (Austin, TX), 2005.

[11] Kennedy, J et al. Particle swarm optimization. Proceedings of ICNN'95 - International Conference on Neural Networks, 4, pages 1942-1948, 1995.

[12] Chuang, L. et al. (2011). Chaotic particle swarm optimization for feature selection in biomedical datasets. Journal of Medical and Biological Engineering, 31(1), 25-34, 2011.

[13] Xue, et al.. Particle swarm optimization for feature selection in classification: A multi-objective approach. IEEE Transactions on Cybernetics, 47(3), 757-770, 2016.

[14] Unler, et al.. A discrete particle swarm optimization method for feature selection in binary classification problems. European Journal of Operational Research, 206(3), 528-539, 2010.

[15] Holland, J. H. Adaptation in natural and artificial systems: An introductory analysis with applications to biology, control, and artificial intelligence. MIT press, 1992.

[16] Dorigo, M., et al. Ant colony system: A cooperative learning approach to the traveling salesman problem. IEEE Transactions on Evolutionary Computation, 1(1), 53-66, 1997.

[17] Omran, M. et al. (2007). Particle swarm optimization method for image clustering. International Journal of Pattern Recognition and Artificial Intelligence, 21(03), 477-498, 2007.

[18] Ghosh, K et al. Medical image segmentation using particle swarm optimization aided clustering techniques. Applied Soft Computing, 26, 174-186, 2015.

[19] Li, X., & Wang, S. (2008). A hybrid particle swarm optimization method for supervised classification. International Journal of Hybrid Intelligent Systems, 5(1), 35-45.

[20] Poli, R., et al. Particle swarm optimization: An overview. Swarm Intelligence, 1(1), 33-57, 2007.

[21] Cui, X et al. Document clustering analysis based on hybrid PSO+ K-means algorithm. Journal of Computer Sciences, 1(1), 27-33, 2005.

[22] Rajesh, R et al. Classification of ECG beats using PSO and decision tree. Procedia Computer Science, 115, 327-336, 2017.

[23] Mehdi, H.et al. Feature selection using particle swarm optimization in text categorization. Journal of Artificial Intelligence and Soft Computing Research (JAISCR), 5(4): 231-238, 2015.