Hybrid Approaches Combining Bio-Inspired Optimization With Machine Learning in Intrusion Detection

Abstract

This chapter aims to delve into the novel integration of machine learning techniques and bio-inspired optimization algorithms within the realm of intrusion detection systems (IDS). In recent decades, IDS have emerged as vital tools for safeguarding data and networks. The detection of intrusions is one of the biggest challenges in network traffic analysis. The proposed research is unique because it aims to develop a system for the detection of intrusions. An ant colony optimization is used in this study to detect intrusions using support vector machines. The proposed system was tested using the Knowledge Discovery and Data Mining (KDD) Cup '99. Dimensionality is a major challenge in network analysis datasets. Dimensionality reduction was achieved using the ant colony optimization algorithm. An ACO approach selects a meaningful subset of features from the full dataset. SVM machine learning algorithms were used to detect intrusions using the selected subset of features. ACO-SVM is therefore more effective at safeguarding a network system from intrusions, based on this analysis.