A Hybrid Swarm Intelligence Algorithm for Intrusion Detection Using Significant Features.

Area of Science:

• Computer Science
• Cybersecurity
• Artificial Intelligence

Background:

• Network security is crucial due to increasing cyberattacks and global internet connectivity.
• Effective intrusion detection systems are vital for protecting information systems.
• Existing methods require optimization for improved accuracy and feature selection.

Purpose of the Study:

• To propose a hybrid algorithm integrating Modified Artificial Bee Colony (MABC) and Enhanced Particle Swarm Optimization (EPSO) for intrusion detection.
• To identify and select the most relevant network traffic features for improved classification.
• To evaluate the effectiveness of the hybrid approach on a benchmark dataset.

Main Methods:

• A hybrid algorithm combining MABC and EPSO was developed for intrusion detection.
• Feature selection was performed to identify critical network traffic patterns.
• The proposed model was evaluated using 10-fold cross-validation on the KDDCup'99 dataset.
• Performance was compared against other machine learning algorithms.

Main Results:

• The hybrid MABC-EPSO algorithm demonstrated superior optimization and classification accuracy.
• Feature selection significantly enhanced the performance of the proposed classification algorithm.
• The method achieved significantly different and improved results compared to existing algorithms.

Conclusions:

• The proposed hybrid MABC-EPSO algorithm offers a promising solution for network intrusion detection.
• Effective feature selection is key to enhancing the accuracy of intrusion detection systems.
• This approach provides a robust method for improving cybersecurity defenses.