Baad: A Self-Optimizing Algorithm for Anomaly Detection
Adeel Shiraz Hashmi1, Tanvir Ahmad2
1Adeel Shiraz Hashmi, Department of Computer Engineering, Jamia Millia Islamia, Delhi, India.
2Tanvir Ahmad, Department of Computer Engineering, Jamia Millia Islamia, Delhi, India.
Manuscript received on 01 May 2019 | Revised Manuscript received on 15 May 2019 | Manuscript published on 30 May 2019 | PP: 499-505 | Volume-8 Issue-7, May 2019 | Retrieval Number: F3754048619/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Anomaly/Outlier detection is the process of finding abnormal data points in datasets or data streams. Anomaly detection finds its application in various fields like network intrusion detection, fraud detection, fault detection, etc. There are many anomaly detection algorithms available in the literature but most of these algorithms require setting of some parameters which significantly affect the performance of the algorithm. These parameters are generally set by hit-and-trial, hence performance is compromised with default or random values. In this paper, the authors propose a self-optimizing algorithm for anomaly detection based on bat meta-heuristic, and named as Bat Algorithm for Anomaly Detection (BAAD). The proposed solution is a non-clustering unsupervised learning approach for anomaly detection. The BAAD algorithm belongs to the category of density-based algorithms, and aims to find the optimal value of neighborhood radius as done in the LOCI (Local Correlation Integral) algorithm for anomaly detection, though the approach used in the proposed solution is different. The algorithm is implemented on Apache Spark for scalability and thus the solution can handle big data as well and provides fast results. Experiments were conducted on various datasets, and the results show that the proposed solution is much accurate than the standard algorithms of anomaly detection.
Keyword: Anomaly Detection, Bat Algorithm, Big Data, Parallel Algorithms.
Scope of the Article: Web Algorithms.