A Hybrid Machine Learning Framework for Anomaly Detection and Electricity Theft Identification in Smart Distribution Networks: A Comprehensive Review

This paper presents a review of hybrid machine learning approaches for electricity theft detection in smart grids using Advanced Metering Infrastructure (AMI) data. The study analyzes supervised, unsupervised, and hybrid anomaly detection techniques including Isolation Forest and Histogram-based Gradient Boosting. Experimental evaluation on a synthetic smart meter dataset demonstrates improved detection accuracy and reduced false positives using hybrid models. The paper also discusses future directions such as Explainable AI (XAI), Federated Learning, and Graph Neural Networks (GNNs) for enhancing smart grid security.