Title :
Circular backpropagation networks embed vector quantization
Author :
Ridella, Sandro ; Rovetta, Stefano ; Zunino, Rodolfo
Author_Institution :
Dept. of Biophys. & Electron. Eng., Genoa Univ., Italy
fDate :
7/1/1999 12:00:00 AM
Abstract :
This letter proves the equivalence between vector quantization (VQ) classifiers and circular backpropagation (CBP) networks. The calibrated prototypes for a VQ schema can be plugged in a CBP feedforward structure having the same number of hidden neurons and featuring the same mapping. The letter describes how to exploit such equivalence by using VQ prototypes to perform a meaningful initialization for BP optimization. The approach effectiveness was tested considering a real classification problem (NIST handwritten digits)
Keywords :
backpropagation; feedforward neural nets; optimisation; pattern classification; vector quantisation; BP optimization; CBP feedforward structure; CBP networks; VQ classifiers; circular backpropagation networks; handwritten digit classification; initialization; vector quantization; Backpropagation; Feedforward neural networks; NIST; Neural networks; Neurons; Nonlinear optics; Prototypes; Switches; Testing; Vector quantization;
Journal_Title :
Neural Networks, IEEE Transactions on
