Online learning vector quantization: A harmonic competition approach basedon conservation network

This paper presents a self-creating neural network in which a conservation principle is incorporated with the competitive learning algorithm to harmon ize equi-probable and equi distortion criteria [1]. Each node is associated with a measure of vitality which is updated after each input presentation. The total amount of vitality in the network at any time is 1, hence the na me conservation. Competitive learning based on a vitality conservation prin ciple is near-optimum, in the sense that problem of trapping in a local min imum is alleviated by adding perturbations to the learning rate during node generation processes. Combined with a procedure that redistributes the lea rning rate variables after generation and removal of nodes, the competitive conservation strategy provides a novel approach to the problem of harmoniz ing equi-error and equi probable criteria. The training process is smooth a nd incremental, it not only achieves the biologically plausible learning pr operty, but also facilitates systematic derivations for training parameters . Comparison studies on learning vector quantization involving stationary a nd nonstationary, structured and nonstructured inputs demonstrate that the proposed network outperforms other competitive networks in terms of quantiz ation error, learning speed, and codeword search efficiency.