Authors: Zakea Il-agure; Belsam Attallah
Addresses: Computer Information Science (CIS), Higher Colleges of Technology (HCT), Abu Dhabi, UAE ' Computer Information Science (CIS), Higher Colleges of Technology (HCT), Abu Dhabi, UAE
Abstract: Anomalies detection is concerned with the problem of finding non-conforming patterns in datasets. Il-agure (2016) described a novel approach to measure the amount of information shared between any random anomaly variables. The CRISP data mining methodology was updated to be applicable for link mining study. The proposed mutual information approach to provide a semantic investigation of the anomalies and the updated methodology can be used in other link mining studies. The purpose of this paper is to evaluate how mutual information interprets semantic anomalies, using density-based cluster technique, via trial 2, which is different than the clustering technique used in trial 1 (hierarchical based cluster). A cluster method allows for many options regarding the algorithm for combining groups, with each choice resulting in a different grouping structure. Therefore, cluster analysis can be an appropriate statistical tool for discovering underlying structures in various kinds of datasets.
Keywords: mutual information; anomalies; DBSCAN algorithm.
International Journal of Grid and Utility Computing, 2019 Vol.10 No.1, pp.36 - 41
Received: 05 Jan 2018
Accepted: 09 Jun 2018
Published online: 24 Dec 2018 *