Article ID Journal Published Year Pages File Type
484904 Procedia Computer Science 2015 8 Pages PDF
Abstract

Data analysis plays amajor role in innovation of new trends in many applications.In most of the current applicationdatabases is being updated day to day. In order to adopt these changes, there is a need to update the present technologies and data mining algorithms in support of changing data. In many of the clustering algorithms the user has to specify the optimum number of clusters prior to execution, for static databases this value remains constant whereas, in the case of dynamic databases the value should be changed. In this paper,we implemented a method to find optimal number of clusters based onfuzzy silhouette on dynamic data by comparing traditional clustering on synthetic data and dynamic customer segmentation.

Related Topics
Physical Sciences and Engineering Computer Science Computer Science (General)