A modified K-Modes clustering algorithm

5Citations
Citations of this article
4Readers
Mendeley users who have this article in their library.

This article is free to access.

Abstract

The main aim of this paper is to apply a new dissimilarity measure, and handle boundary data properly in K-Modes clustering thereby increasing the clustering efficiency. Moreover our proposed algorithm identifies the outlier data efficiently. © Springer-Verlag 2013.

Cite

CITATION STYLE

APA

Bishnu, P. S., & Bhattacherjee, V. (2013). A modified K-Modes clustering algorithm. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8251 LNCS, pp. 60–66). https://doi.org/10.1007/978-3-642-45062-4_7

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free