AZE

“Weighted Clustering for Anomaly Detection in Big Data” article published

2394
A A A
July 06 2018

“Weighted Clustering for Anomaly Detection in Big Data” (DOI: 10.19139/soic.v6i2.404) article , co-authored by Academician Rasim Alguliev (academician-secretary of ANAS, Chief of Information Technologies Institute) , Ramiz Aliguliyev (corresponding member of ANAS, PhD of technical sciences), Yadigar Imamverdiyev (chief of department, PhD of technical sciences) and Lyudmila Suxostat (senior scientific worker, PhD of technical sciences) published by the "Statistics, Optimization & Information Computing" journal, which is published by the US-based “International Academic Press”.

The concept of big data is intended to work with information that is frequently updated with large volumes and changeable content. It is crucial to choose the appropriate mechanism for detecting anomalies in the Big Data. The article suggests an algorithm based on weighted clustering to detect anomalies in real Big Data. The clusters are obtained the weight gains by summing the weight coefficients of each point of the data set. Comparative analysis of the method proposed by the K-Means method was carried out on a large database. The results of the experiments have shown the effectiveness of the proposed approach in both clusterization and detection of anomalies.

The article is prepared within the framework of the “Development of methods and algorithms for ensuring information security in Big Data" project, funded by the Science Development Foundation under the President of the Republic of Azerbaijan. (Grant № EİF-KETPL-2-2015-1(25) - 56/05/1).

“Statistics, Optimization & Information Computing” is indexed in “Scopus” (Elsevier), “Crossref”, “Google Scholar”, DOAJ and other international science markets.

Source: Science.gov.az

WISE WORDS