Self-organization and associative memory: 3rd edition
Self-organization and associative memory: 3rd edition
Competitive learning algorithms for vector quantization
Neural Networks
Representation-burden Conservation Network Applied to Learning VQ (NPL270)
Neural Processing Letters
Harmonic competition: a self-organizing multiple criteria optimization
IEEE Transactions on Neural Networks
Codeword distribution for frequency sensitive competitive learning with one-dimensional input data
IEEE Transactions on Neural Networks
Hi-index | 0.00 |
In a recent publication [1], it was shown that abiologically plausible RCN (Representation-burdenConservation Network) in which conservation isachieved by bounding the summed representation-burdenof all neurons at constant 1, is effective in learningstationary vector quantization. Based on theconservation principle, a new approach for designinga dynamic RCN for processing both stationary andnon-stationary inputs is introduced in thispaper. We show that, in response to the inputstatistics changes, dynamic RCN improves itsoriginal counterpart in incremental learningcapability as well as in self-organizing the networkstructure. Performance comparisons between dynamic RCN and other self-development models arealso presented. Simulation results show that dynamic RCN is very effective in training anear-optimal vector quantizer in that it manages tokeep a balance between the equiprobable andequidistortion criterion.