An Influential Algorithm for Outlier Detection

The Subject of Data Mining, which is very vast in the direction of data analyzing , data preprocessing, data
extracting, knowledge data discovery, extracting hidden data and many more. Here we discuss one of the
parts of the data mining. Which is clustering, in the clustering there is the major issue to eradicate outliers
from the data. Here we do work for detect outliers from the data sets.
In the paper we had detailed studied about the different-different ways for detecting outliers. Likewise cluster
based, distance based, density based etc. this approaches are used in the different-different methods of outlier
detection (PAM, CLARA, CLARANCE, ECLARA). Here we choose the best one for further enhancement on
the basis of their performance. new hybrid algorithm proposed with the classical clustering method for the
improving performance on the basis of accuracy, error rate, time complexity.