Interpretable Anomaly Detection: A Hybrid Approach Using Rule-Based and Machine Learning Techniques

Tools

Ouarbya, Lahcen and Rahul, Mohite. 2023. 'Interpretable Anomaly Detection: A Hybrid Approach Using Rule-Based and Machine Learning Techniques'. In: 2024 IEEE 9th International conference for Convergence in Technology (I2CT). Vivanta Pune, Hinjawadi, Hinjawadi Road Hinjawadi Village, Hinjawadi, Pune, India 5 - 7 April 2024. [Conference or Workshop Item] (Forthcoming)

Preview

Text (Interpretable_Anomaly_Detection__A_Hybrid_Approach_Using_Rule_Based_and_Machine_Learning_Techniques.pdf)
Interpretable_Anomaly_Detection__A_Hybrid_Approach_Using_Rule_Based_and_Machine_Learning_Techniques.pdf - Accepted Version
Download (81kB) | Preview

Abstract or Description

Anomaly detection is a critical aspect of ensuring the security and reliability of various systems in diverse domains, including cybersecurity, finance, and industrial processes. Traditional black-box anomaly detection methods often lack interpretability, making it challenging for users to understand the reasoning behind the detection of anomalies. In this research, we propose a novel hybrid approach that combines rule-based and machine learning techniques to enhance the interpretability of anomaly detection systems. Our method integrates a rule-based system that generates interpretable anomaly detection rules with machine learning components that leverage complex pattern recognition and classification capabilities. We evaluate the proposed approach on a diverse set of real-world datasets, demonstrating its effectiveness in identifying anomalies while providing transparent explanations of the detection process. Through comprehensive experimentation and comparative analysis with existing state-of-the-art methods, we showcase the superior interpretability and performance of our hybrid approach. Our findings highlight the significance of interpretability in anomaly detection systems and underscore the potential of the proposed approach for enhancing transparency and trust in critical decision-making processes. This research contributes to the advancement of interpretable anomaly detection techniques and opens avenues for future research in the domain of transparent and reliable anomaly detection systems.

Item Type:

Conference or Workshop Item (Paper)

Additional Information:

“© 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.”

Keywords:

Anomaly detection, Outlier analysis, Interpretability, Rule-based systems, Machine learning, Hybrid approach, Trans- parency, Trustworthy AI, Pattern recognition, Classification, Data analysis, Cybersecurity, Financial fraud detection, Industrial pro- cesses, Explainable AI

Related URLs: