A Comprehensive Review and Analysis of Anomaly Detection

Abstract

This research work emphasizes on the challenges and issues faced by a researcher while working on anomaly detection using deep learning. The motivation behind this research is to highlight the initial challenges and complexities encountered at the commencement of anomaly detection. The primary hurdle involves the precise identification and categorization of anomalies, with this paper expounding on various anomaly types and their distinctive characteristics. Another challenge arises in the analytical process of discerning and selecting the most optimal model, considering their varying levels of accuracy. So to make this task simpler the research elucidates various deep learning models. Subsequently, it conducts a comprehensive review of the work undertaken by different researchers in the realm of anomaly detection, comparing their learnings and outputs. And the most accurate model is suggested [37].