Volume 22, Issue 4 (12-2025)
jor 2025, 22(4): 1-15 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Refahi Sheikhani A H. A Suitable Optimization Algorithm for the Feature Selection Problem. jor 2025; 22 (4) :1-15
URL: http://jamlu.liau.ac.ir/article-1-2286-en.html
URL: http://jamlu.liau.ac.ir/article-1-2286-en.html
Department of Applied Mathematics, Faculty of Mathematical Sciences, Lahijan Branch, Islamic AzadUniversity, Lahijan, Iran , ah_refahi@yahoo.com
Abstract: (41 Views)
The classification operation and achieving lower classification error, which leads to higher accuracy, do not require using all the features of a dataset. By selecting meaningful features and reducing the dimensionality of the feature vector, the performance of the proposed method can be improved. Our experimental results show that selecting specific subsets of features can lead to better performance on different datasets. Specifically, in the GISETTE dataset, selecting 20 features through the ACO algorithm achieved an accuracy of 98.18%. Also, in other datasets such as RELATHE and PROSTATE-GE, the accuracy was 77.85% and 97.77%, respectively. The process of finding the appropriate number of features is usually time-consuming. Hence, we consider the feature selection problem as an optimization problem and leave its solution to meta-heuristic methods. In this paper, an approach for feature selection using the Ant Colony Optimization (ACO) algorithm is presented. The two-layer perceptron as a classifier uses the features selected by the ACO algorithm in the classification operation. To evaluate the ACO method, a new and sparse norm has been used, which we have named the method ANT-ANN-SSN. The experimental results show that the ANT-ANN-SSN method has performed better than other methods in all data sets.
Type of Study: Research |
Subject:
Special
Received: 2024/10/29 | Accepted: 2025/01/21 | Published: 2025/12/22
Received: 2024/10/29 | Accepted: 2025/01/21 | Published: 2025/12/22
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
