Please use this identifier to cite or link to this item:
Title: Relative Dimensionality of Feature Selection Based on Collective Behavior in Cluster Databasesd
Authors: Veeraiah, D
Vasumathi, D
Keywords: Social Networking;Collective Behavior;Classification with Categorical Data;Relational Databases
Issue Date: Aug-2017
Publisher: NISCAIR-CSIR, India
Abstract: Problem of clustering categorical data in social streams via link based cluster ensemble is being competitive with conventional algorithms. By observing these procedures conventional algorithms produce last information divided based on partial information. It is observed that individual behavior of each feature present in categorical data then cluster ensemble approach may fail due to insufficient analysis of feature selection in public web sites. Big amount of information producing day by day in public networks like Face book, Twitter, and YouTube present possibilities, difficulties peruse aggregate activities on a broad. Available goal is to grasp to estimate combined actions in public networking. In this system it is proposed to use an edge-centric grouping plan to get rare public measurements. With these rare public measurements, the proposed way can effectively manage systems of an incredible number of stars during indicating a similar forecast representation to other techniques of non-scalable. Our experimental results show efficient dimensionality selection in categorical data.
Page(s): 468-472
ISSN: 0975-1084 (Online); 0022-4456 (Print)
Appears in Collections:JSIR Vol.76(08) [August 2017]

Files in This Item:
File Description SizeFormat 
JSIR 76(8) 468-472.pdf147.72 kBAdobe PDFView/Open

Items in NOPR are protected by copyright, with all rights reserved, unless otherwise indicated.