An Incremental Majority Voting Approach for Intrusion Detection System Based on Machine Learning

Abstract

With the rapid growth of internet technologies and network-based applications, cyber threats and security attacks have significantly increased. Intrusion Detection Systems (IDS) play a critical role in identifying unauthorized access, malicious activities, and network vulnerabilities. Traditional IDS approaches often suffer from high false alarm rates and limited detection accuracy when dealing with large-scale and complex network traffic. To address these challenges, this research proposes an Incremental Majority Voting based Intrusion Detection System using machine learning techniques. The proposed model combines multiple classifiers and applies a majority voting mechanism to improve detection performance. The incremental learning approach enables the system to update its knowledge dynamically with new data, allowing it to adapt to evolving cyber threats. Experimental evaluation demonstrates that the proposed model improves detection accuracy while reducing false positives. The system is capable of identifying different types of network attacks including DoS, Probe, R2L, and U2R attacks. The results indicate that the incremental majority voting model provides better performance compared to individual machine learning classifiers. This approach can be effectively applied in modern network security systems to enhance intrusion detection efficiency.