A global k-means approach for autonomous cluster initialization of probabilistic neural network

This paper focuses on the statistical based Probabilistic Neural Network (PNN) for pattern classification problems with Expectation â�� Maximization (EM) chosen as the training algorithm. This brings about the problem of random initialization, which means, the user has to predefine the number of clu...

Full description

Bibliographic Details
Main Authors:	Chang, R.K.Y., Loo, C.K., Rao, M.V.C.
Format:	Article
Published:	2008
Subjects:	T Technology (General)
Online Access:	http://en.scientificcommons.org/55706914 http://en.scientificcommons.org/55706914 http://eprints.um.edu.my/5158/1/A_Global_k%2Dmeans_approach_for_autonomous_cluster_initialization_of_probabilistic_neural_network.pdf

id	um-5158
recordtype	eprints
spelling	um-51582013-03-19T00:15:14Z A global k-means approach for autonomous cluster initialization of probabilistic neural network Chang, R.K.Y. Loo, C.K. Rao, M.V.C. T Technology (General) This paper focuses on the statistical based Probabilistic Neural Network (PNN) for pattern classification problems with Expectation â�� Maximization (EM) chosen as the training algorithm. This brings about the problem of random initialization, which means, the user has to predefine the number of clusters through trial and error. Global k-means is used to solve this and to provide a deterministic number of clusters using a selection criterion. On top of that, Fast Global k-means was tested as a substitute for Global k-means, to reduce the computational time taken. Tests were done on both homescedastic and heteroscedastic PNNs using benchmark medical datasets and also vibration data obtained from a U.S. Navy CH-46E helicopter aft gearbox (Westland) 2008 Article PeerReviewed application/pdf http://eprints.um.edu.my/5158/1/A_Global_k%2Dmeans_approach_for_autonomous_cluster_initialization_of_probabilistic_neural_network.pdf http://en.scientificcommons.org/55706914 Chang, R.K.Y.; Loo, C.K.; Rao, M.V.C. (2008) A global k-means approach for autonomous cluster initialization of probabilistic neural network. Informatica <http://eprints.um.edu.my/view/publication/Informatica.html>, 32. pp. 219-225. ISSN 0350-5596 http://eprints.um.edu.my/5158/
repository_type	Digital Repository
institution_category	Local University
institution	University Malaya
building	UM Research Repository
collection	Online Access
topic	T Technology (General)
spellingShingle	T Technology (General) Chang, R.K.Y. Loo, C.K. Rao, M.V.C. A global k-means approach for autonomous cluster initialization of probabilistic neural network
description	This paper focuses on the statistical based Probabilistic Neural Network (PNN) for pattern classification problems with Expectation â�� Maximization (EM) chosen as the training algorithm. This brings about the problem of random initialization, which means, the user has to predefine the number of clusters through trial and error. Global k-means is used to solve this and to provide a deterministic number of clusters using a selection criterion. On top of that, Fast Global k-means was tested as a substitute for Global k-means, to reduce the computational time taken. Tests were done on both homescedastic and heteroscedastic PNNs using benchmark medical datasets and also vibration data obtained from a U.S. Navy CH-46E helicopter aft gearbox (Westland)
format	Article
author	Chang, R.K.Y. Loo, C.K. Rao, M.V.C.
author_facet	Chang, R.K.Y. Loo, C.K. Rao, M.V.C.
author_sort	Chang, R.K.Y.
title	A global k-means approach for autonomous cluster initialization of probabilistic neural network
title_short	A global k-means approach for autonomous cluster initialization of probabilistic neural network
title_full	A global k-means approach for autonomous cluster initialization of probabilistic neural network
title_fullStr	A global k-means approach for autonomous cluster initialization of probabilistic neural network
title_full_unstemmed	A global k-means approach for autonomous cluster initialization of probabilistic neural network
title_sort	global k-means approach for autonomous cluster initialization of probabilistic neural network
publishDate	2008
url	http://en.scientificcommons.org/55706914 http://en.scientificcommons.org/55706914 http://eprints.um.edu.my/5158/1/A_Global_k%2Dmeans_approach_for_autonomous_cluster_initialization_of_probabilistic_neural_network.pdf
first_indexed	2018-09-05T16:49:54Z
last_indexed	2018-09-05T16:49:54Z
_version_	1610833543001276416

A global k-means approach for autonomous cluster initialization of probabilistic neural network

Similar Items