|
iRepository at Perpustakaan UniMAP >
UNIVERSITY LIBRARY >
Conference Papers >
Please use this identifier to cite or link to this item:
http://hdl.handle.net/123456789/10330
|
| Title: | Kernel-Induced Bubble Agglomeration Algorithm for unsupervised classification: An improved clustering methodology without prior information |
| Authors: | Lim, Eng Aik |
| E-mail: | ealim@unimap.edu.my |
| Keywords: | Kernel function
Bubble Agglomeration
Euclidean distance
Data classification
Similarity measure
Regional Conference on Applied and Engineering Mathematics (RCAEM) |
| Issue Date: | 2-Jun-2010 |
| Publisher: | Universiti Malaysia Perlis (UniMAP) |
| Department: | Institut Matematik Kejuruteraan |
| Citation: | Vol.4(8), p.395-400 |
| Series/Report no.: | Proceedings of the 1st Regional Conference on Applied and Engineering Mathematics (RCAEM-I) 2010 |
| Abstract: | This paper introduces an improved unsupervised clustering
algorithm, named Kernel-Induced Bubble Agglomeration.
In this paper, the conventional Bubble Agglomeration
algorithm is extended by calculating the Euclidean distance
of each data point based on a kernel-induced distance
instead of the conventional sum-of-squares distance. The
kernel function is a generalization of the distance metric
that measures the distance between two data points as the
data points are mapped into a high dimensional space. By
using a kernel function, data that are not easily separable in
the original space can be clustered into homogeneous
groups in the implicitly transformed high dimensional
feature space. Application of the conventional Bubble
Agglomeration algorithm and the Kernel-induced Bubble
Agglomeration algorithm to well-known data sets showed
the superiority of the proposed approach. |
| Description: | 1st Regional Conference on Applied and Engineering Mathematics (RCAEM-I) 2010 organized by Universiti Malaysia Perlis (UniMAP) and co-organized by Universiti Sains Malaysia (USM) & Universiti Kebangsaan Malaysia (UKM), 2nd - 3rd June 2010 at Eastern & Oriental Hotel, Penang. |
| URI: | http://hdl.handle.net/123456789/10330 |
| Appears in Collections: | Conference Papers
|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.
|
