Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{184990,
        author = {Ganeshh Rathod},
        title = {Optimizing Feature Selection in Intrusion Detection Using Fisher Score Algorithm: An Analytical Study},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {4},
        pages = {4556-4569},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=184990},
        abstract = {— Traditional intrusion detection methods often fail to meet the unique security requirements of IoT applications, raising serious security concerns due to the rapid expansion of the Internet of Things (IoT) -connectivity and its extensive property, complex structure, and complexity. Machine learning algorithms for intrusion detection have emerged as a solution, offering promise in protecting these complex systems. This paper examines a comprehensive analysis of an Intrusion Detection System (IDS) in an IoT system. Recognizing the important role of feature extraction in IDS, this study aims to help researchers by providing insights on data set selection and confirming the effectiveness of the Fisher Score algorithm. Through a careful comparative analysis using established selection methods—Mutual Information, Chi-Square, Principal Component Analysis (PCA), and Recursive Feature Elimination (RFE)— this study seeks to help researchers choose the most appropriate feature extraction technique in intrusion detection within the IoT framework. Using logistic regression as a classification model, this study allows a thorough analysis and comparison of different selection methods. The results highlight the importance and accuracy of the Fisher Score algorithm in selecting key features for Intrusion Detection in IoT systems. It is worth mentioning, that this research is limited to a specific dataset, NBaIot, which is considered the best and most up-to-date for Intrusion Detection. The specificity observed in this study is dependent on the number of items and data sets. Although the findings may be variable, the selected sets and item sizes have informed the identification of interventions in this study that are deemed optimal.},
        keywords = {Machine Learning, Intrusion Detection System, Fisher Score, Mutual Information, Chi-Square, Principal Component Analysis (PCA), Recursive Feature Elimination (RFE).},
        month = {October},
        }