Thursday 22nd of February 2018

Implementation of DB-Scan in Multi-Type Feature CoSelection for Clustering

K.Parimala and Dr. V.Palanisamy

Feature Selection is a preprocessing technique in supervised learning for improving predictive accuracy while reducing dimension in clustering and categorization. Multitype Feature Coselection for Clustering (MFCC) with hard k-means is the algorithm which uses intermediate results in one type of feature space enhancing feature selection in other spaces, better feature set is co-selected by heterogeneous features to produce better cluster in each space. Db-Scan is a density-based clustering algorithm finding a number of clusters starting from the estimated density distribution of corresponding nodes. It is one of the most common clustering algorithms and also most cited in scientific literature, as a generalization of DBSCAN to multiple ranges, effectively replacing the parameter with a maximum search radius.This paper presents the empirical results of the MFCC algorithm with Db-scan and also gives the comparison results of MFCC with hard k-means and DB-Scan. DB-Scan clustering is proposed for getting the quality clustering against the outliers and time criteria is less than any other clustering in high density data set.

Keywords: Feature Selection, MFCC, Db-Scan.

Download Full-Text


Mrs.K.Parimala.,MCA, Research Scholar in Computer Science in Alagappa University, Karaikudi, INDIA, under the guidance of Dr.V.Palanisamy, Professor & Head In-Charge, Department of Computer Science & Engineering, Alagappa University. Currently working as Assistant Professor, in NMS SVN College with a teaching experience of 14 years.

Dr. V.Palanisamy
Dr. V.PalaniSamy, MCA, MTech(Adv.IT), Ph.D, Professor & Head In-Charge, Department of Computer Science & Engineering, Alagappa University, Karaikudi, TamilNadu, INDIA, specialized in Algorithms, Wireless Networks & Network Security. He has 20 years of teaching experience and 15 years of Research Experience.

IJCSI Published Papers Indexed By:





IJCSI is a refereed open access international journal for scientific papers dealing in all areas of computer science research...

Learn more »
Join Us

Read the most frequently asked questions about IJCSI.

Frequently Asked Questions (FAQs) »
Get in touch

Phone: +230 911 5482

More contact details »