Title :
Clustering of Gaussian probability density functions using centroid neural network
Author :
Park, Dong-Chul ; Kwon, Oh-Hyun ; Suk, Minsoo
Author_Institution :
Dept. of Inf. Eng., Myong Ji Univ., KyungKi-Do, South Korea
fDate :
2/20/2003 12:00:00 AM
Abstract :
An unsupervised competitive neural network algorithm for clustering mixtures of Gaussian probability density functions is proposed. The algorithm based on centroid neural network with Bhattacharyya distance is evaluated in the context of speech recognition and the results show that it can reduce the Gaussian mixtures by almost 60% over the k-means algorithm.
Keywords :
Gaussian distribution; hidden Markov models; neural nets; pattern clustering; speech recognition; unsupervised learning; Bhattacharyya distance; Gaussian probability density function; centroid neural network; clustering analysis; continuous density hidden Markov model; speech recognition; unsupervised competitive learning algorithm;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:20030247
