Title :
A competitive and selective learning method for designing optimal vector quantizers
Author :
Ueda, Naonori ; Nakano, Ryohei
Author_Institution :
NTT Commun. Sci. Lab., Kyoto, Japan
Abstract :
A new competitive learning method with a `selection´ mechanism is proposed for the design of optimal vector quantizers. A basic principle called the `equidistortion principle´ for designing optimal quantizers is derived theoretically, and a new learning algorithm based on this principle is presented. Unlike conventional algorithms based on the `conscience´ mechanism, the proposed algorithm can minimize distortion without a particular initialization procedure, even when the input data cluster in a number of regions in the input vector space. The performance of this method is compared with that of the conscience learning method
Keywords :
learning (artificial intelligence); neural nets; vector quantisation; competitive learning; conscience learning method; input data cluster; input vector space; learning algorithm; optimal vector quantizers; selective learning method; Algorithm design and analysis; Clustering algorithms; Design methodology; Distortion measurement; Image coding; Laboratories; Learning systems; Neurons; Rate distortion theory; Vector quantization;
Conference_Titel :
Neural Networks, 1993., IEEE International Conference on
Conference_Location :
San Francisco, CA
Print_ISBN :
0-7803-0999-5
DOI :
10.1109/ICNN.1993.298769
