Models & Optimisation and Mathematical Analysis Journal
Volume 10, Numéro 1, Pages 20-25
2022-12-31
Intrusion Detection System Based M-best Features Selection In Manet
Authors : Khalladi Rachid . Rebbah Mohammed . Smail Omar .
Abstract
MANETs networks are mobile units interconnected between them without infrastructures which make them vulnerable to different types of attacks. Although several techniques have been proposed as an endeavour to remedy this issue, they are still insufficient. In this work, a technique based on machine learning, more precisely on the random forest algorithm with the selection of the best features, is proposed. The latter is tested on the NSL-KDD dataset. The results found were very satisfying in terms of Accuracy 99,625%, Precision 99,85 %, Recall 99,83 % and F1-Score 99,84%. Thus, the results have improved when compared with those of other techniques.
Keywords
Machine learning ; Intrusion detection ; NSL-KDD ; Random Forest ; Most Best Features Selection
Les articles similaires
Dernouni Fouzia
.
Salah Boualem
.
Kettab Ahmed
.
pages 75-83.
Zaoui Mohamed
.
Himouri Slimane
.
Kadri Tahar
.
Benaouina Charef
.
pages 307-324.
Guettaf Abdelkader
.
Djkhioua Tahir
.
Abid Farid Zakaria
.
pages 188-206.
Benyahia Samia
.
Meftah Boudjelal
.
pages 23-27.
