An object whose behaviour is found to be different from others in a dataset is said to be an outlier. The existing outlier detection algorithms are able to detect outliers only in static datasets, but are found to be inappropriate, when it comes to dynamic datasets where data arrive continuously in a stream-lined fashion viz., sensor data. To deal with such steam data, Association rule mining serves as a best technique, where frequent item sets are internally evaluated from the data, in an iterative fashion. Outlier detection techniques for static datasets include cluster analysis, where clusters are being generated from the data using k-Means clustering to discover outliers. In this paper, we propose two different approaches for outlier detection. One uses association rule based technique on dynamic datasets and the other uses K-means clustering and distance based approach on static datasets to prune local outliers. Experiments are conducted on different variants of static and dynamic datasets to detect the deviant objects (outliers) effectively in fewer computations.

outlier, static data, dynamic data, association rule mining, cluster analysis

[1] Li-Jen Kao, Yo-Ping Huang, “Association rules based algorithm for identifying outlier transactions in data stream,” IEEE International Conference on Systems, Man, and Cybernetics, Oct. 14-17, 2012.
[2] J.H. Chang and W.S. Lee, “Finding recent frequent item sets adaptively over online data streams,” in Proceedings of the 9th ACM SIGKDD, Washington, DC, USA, pp.487-492, August 2003.
[3] E.M. Knorr and R.T. Ng, “Algorithms for mining distance-based outliers in large databases,” In Proceedings 24th International Conference on Very Large Data Bases, VLDB, pp. 392-403, 1998.
[4] P. Rajendra, D. Jatindra Kumar, N. Sukumar, “An outlier detection method based on clustering,” International Conference on Emerging Applications of Information Technology, 2011.
[5] F. Angiulli, S. Basta, and C. Pizzuti, “Distance-based detection and prediction of outliers,” IEEE Transactions on Knowledge and Data Engineering, 18:145-160, 2006.
[6] F. Angiulli and C. Pizzuti, “Fast outlier detection in high dimensional spaces,” In PKDD ’02: Proceedings of the 6th European Conference on Principles of Data Mining and Knowledge Discovery, pp. 15-26, 2002.
[7] F. Angiulli and C. Pizzuti, “Outlier mining in large high-dimensional data sets,” IEEE Transactions on Knowledge and Data Engineering, 17:203-215, 2005.
[8] M.M. Breunig, H.-P. Kriegel, R.T. Ng, and J. Sander, “LOF: identifying density-based local outliers,” SIGMOD Rec., 29(2):93-104, 2000.
[9] K. Zhang, M. Hutter, and H. Jin, “A new local distance-based outlier detection approach for scattered real-world data,” In PAKDD ’09: Proceedings of the 13th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, pp.813-822, 2009.
[10] K. Narita and H. Katigawa, “Outlier detection for transactional Databases using association rules,” in Proceedings of the 9th International Conference on Web-Age Information Management, Zhangjiajie, Hunan, pp. 373-380, July 2008.
[11] R.S. Walse, G.D. Kurundkar, P.U. Bhalchandra, “A Review: Design and Development of Novel Techniques for Clustering and Classification of Data”, International Journal of Scientific Research in Computer Sciences and Engineering, Vol. 06, pp. 19-22, Jan-2018.
[12] Namrata Ghuse, Pranali Pawar, Amol Potgantwar, “An Improved Approach For Fraud Detection in Health Insurance Using Data Mining Techniques”, International Journal of Scientific Research in Network Security and Communication, vol. 5, issue 5, June-2017.