Comparative Performance Analysis of Machine Learning Algorithms: Random Cut Forest, Robust Random Cut Forest, and Amazon Sage Maker Random Cut Forest for Intrusion Detection Systems Using the CIS IDS 2017 Dataset
Senthilkumar Perumal, Kumaresan Devarajan · Turkish Journal of Engineering · 2025
Dynamic cyber threats are screaming for better anomaly detection techniques in Intrusion Detection Systems. Organizations today are hugely dependent on digital infrastructures for which effective security is priceless. The following research article does a critical and comparative analysis among three popular algorithms, namely Amazon Sage Maker Random Cut Forest, Robust Random Cut Forest, and traditional Random Cut Forest. Using the CIS IDS 2017 dataset with multifaceted network traffic features together with the labeled type of attack, this work rigorously tests the performance in anomaly detection that may show potential intrusion, robustness, scalability, and adaptability of each algorithm. The comparative analysis does the performance metrics of each algorithm based on accuracy, precision, recall, and F1-score in a real-world setting. The findings are expected to provide useful insights toward optimizing IDS frameworks for hi-tech cybersecurity resilience. Finally, an organization can make decisions on its strategy regarding cyber security by being enlightened on the strengths and weaknesses of algorithms. In essence, this paper contributes to the larger body of research on enhancing intrusion detection methodologies in an environment that is confronted by sophisticated cyber-attacks.