A Survey On: Distance Based Outlier Detection

Smita Patil, P. D.Chouksey

Abstract: Data mining is technique & used for outlier Detection. Outlier is a data point which is different from the rest of data. Outlier Detection finds the pattern which is not similar to regular behavior. The entire methodologies for outlier detection can be broadly categorized as supervised outlier detection methods, semi-supervised outlier detection methods and unsupervised outlier detection methods. Unsupervised outlier detection methods have been proved to be prominent in most cases, where high dimensional data come in practice. Outlier detection can usually be considered as a pre-processing step for locating, in a data set, those objects that do not conform to well defined notions of expected behavior. In the previous work, Antihub method and unsupervised method is used for outlier detection. Here the distance based outlier detection is proposed by using antihub and semisupervised learning.

Keywords: Outlier, K-NN, High dimensional dataset, Hubness, Antihub

How to Cite?: Smita Patil, P. D.Chouksey, "A Survey On: Distance Based Outlier Detection", Volume 5 Issue 1, January 2016, International Journal of Science and Research (IJSR), Pages: 280-282, https://www.ijsr.net/getabstract.php?paperid=NOV152729, DOI: https://dx.doi.org/10.21275/NOV152729