Robust machine learning based Intrusion detection system using simple statistical techniques in feature selection

Sunil Kaushik ¹

Akashdeep Bhardwaj ²

Ahmad Almogren ³

Salil Bharany ⁴

Ayman Altameem ⁵

Ateeq Ur Rehman ⁶

Seada Hussen ⁷

Habib Hamam ^{8, 9, 10, 11}

Hide authors affiliations

American Towers (ATC TIPL), Gurgaon, India

Center of Excellence (Cybersecurity), School of Computer Science, UPES, Dehradun, India

Department of Computer Science, College of Computer and Information Sciences, King Saud University, Riyadh, Saudi Arabia |

⁴

Chitkara University Institute of Engineering and Technology, Chitkara University, Mohali, India |

⁵

Department of Natural and Engineering Sciences, College of Applied Studies and Community Services, King Saud University, Riyadh, Saudi Arabia |

⁶

School of Computing, Gachon University, Seongnam-si, Republic of Korea |

⁷

Department of Electrical Power, Adama Science and Technology University, Adama, Ethiopia |

⁸

Faculty of Engineering, Uni de Moncton, Moncton, Canada |

⁹

School of Electrical Engineering, University of Johannesburg, Johannesburg, South Africa |

¹⁰

International Institute of Technology and Management (IITG), Libreville, Gabon |

¹¹

Bridges for Academic Excellence - Spectrum, Tunis, Tunisia

Publication type: Journal Article

Publication date: 2025-02-01

Springer Nature

Journal: Scientific Reports

scimago Q1

SJR: 0.900

CiteScore: 7.5

Impact factor: 3.8

ISSN: 20452322

DOI: 10.1038/s41598-025-88286-9

Copy DOI

Abstract

There are serious security issues with the quick growth of IoT devices, which are increasingly essential to Industry 4.0. These gadgets frequently function in challenging environments with little energy and processing power, leaving them open to cyberattacks and making it more difficult to implement intrusion detection systems (IDS) that work. In order to address this issue, this study presents a unique feature selection algorithm based on basic statistical methods and a lightweight intrusion detection system. This methodology improves performance and cuts training time by 27–63% for a variety of classifiers. By utilizing the most discriminative features, the suggested methods lower the computational overhead and improve the detection accuracy. The IDS achieved over 99.9% accuracy, precision, recall, and F1-Score on the dataset IoTID20, with consistent performance on the NSLKDD dataset.