neural network competitive learning algorithms are widely used for vector quantization. In this paper, some typical competitivelearningalgorithms have been specially investigated, analyzed and their performances hav...
详细信息
ISBN:
(纸本)0780342534
neural network competitive learning algorithms are widely used for vector quantization. In this paper, some typical competitivelearningalgorithms have been specially investigated, analyzed and their performances have also been evaluated. A new competitivelearning algorithm based on the neuron winning probability is presented for vector quantization. Unlike the traditional competitivelearningalgorithms where only one neuron will win and learn in each competition, every neuron in the proposed probability sensitive competitivelearning algorithm (PSCL) will win to some extent, depending on its winning probability and adjustment of distortion distance to the input vector. The new algorithm is shown to be efficient to overcome the problem of neuron underutilization.
暂无评论