Anomaly detection refers to the process of identifying
anomalies or distinct patterns in data that might potentially point to issues.
The motivation to study various anomaly detection techniques lies in the
critical need to identify and respond to unusual patterns or deviations across
multiple domains. There are various methods like statistical methods that are
effective for data with known distributions, proximity-based and
clustering-based methods excel in spatial data analysis, etc. Graph-based
methods detect anomalies by disrupted connectivity patterns to enhance
robustness and performance in diverse applications. In this article, some
graph-based anomaly detection techniques have been studied in a nutshell and
also some future directions to improve the technique of detecting anomalies in
data have been given.
Author(s) Details
Debajit Sensarma
Department of Computer Science, Vivekananda Mission
Mahavidyalaya, Purba Medinipur, Haldia, India.
Please see the link:- https://doi.org/10.9734/bpi/strufp/v7/1068
No comments:
Post a Comment