Data Mining: Finding Outliers from Different Types of Data using Dissimilarity Data Structure

An Outlier is an extreme value in a data set. Using clustering techniques we can detect outliers. Outlier means values that are far away from any cluster. In this paper we tried to find out outliers from Inter-Scaled Variables, Binary Variables, Categorical and Ordinal Variables by using Dissimilarity Data Structure. All similar objects are grouped and objects which are not belonging into any cluster are considered as outliers.