Metaheuristics Based Dimensionality Reduction With Deep Learning Driven False Data Injection Attack Detection For Enhanced Network Security

This article discusses the use of a new hybrid metaheuristics-based dimensionality reduction with deep learning technique for detecting unobservable false data injection attacks in distribution systems. The technique utilizes a combination of data normalization, feature selection, and stacked autoencoder methods to improve detection outcomes. The study highlights the importance of addressing cyberattacks in smart grid systems and presents a potential solution for enhanced network security.