A Feature Selection Algorithm for Intrusion Detection System Based on Binary PSO

When developing models using machine learning, feature selection is an extremely important step. The process of selecting features is an essential part of the construction of an intrusion detection system (IDS). We present an IDS that is based on deep convolutional neural networks (DCNN). Two convol...

Full description

Saved in:
Bibliographic Details
Published in2023 Intelligent Computing and Control for Engineering and Business Systems (ICCEBS) pp. 1 - 10
Main Authors Vengaimarbhan, D., Srinivas, B., Praveena, MS. G.
Format Conference Proceeding
LanguageEnglish
Published IEEE 14.12.2023
Subjects
Classification algorithms
Convolutional neural networks
deep learning
Feature extraction
feature selection
intrusion detection
Network intrusion detection
Optimization
Training
whale optimization
Whale optimization algorithms
Online AccessGet full text

Cover

More Information
Summary:When developing models using machine learning, feature selection is an extremely important step. The process of selecting features is an essential part of the construction of an intrusion detection system (IDS). We present an IDS that is based on deep convolutional neural networks (DCNN). Two convolutional layers and three fully linked dense layers make up a deep convolutional neural network (DCNN). The approach that has been suggested tries to increase performance while also reducing the amount of processing resources required. Experiments were carried out using the IoTID20 dataset serving as the basis. A number of different measures, including accuracy, precision, recall, and F1-score, were used in the process of evaluating the suggested model's performance. whale optimization is used, in which Adam, AdaMax, and Nadam's performance was at its highest possible level. The results of all of the experimental analyses suggest that the suggested method has 97% accuracy.
DOI:10.1109/ICCEBS58601.2023.10448587