Adaptive Learning Algorithm in Tree-structured Self-organizing Feature Map
Takashi Yamaguchi, Takumi Ichimura, Kenneth James Mackin · Institutional Repositories DataBase (IRDB) · 2010
Self-Organizing Feature Map is a layered neural network consisting of an input layer and a competitive layer for the data visualization and vector quantization.The accuracy of SOM vector quantization depends on the number of competitive layer's neurons.Therefore, when an unknown data set is given, it is difficult to decide the sufficient competitive layer size.In this paper, we propose a hierarchical competitive layer adaptation method in order to find out the sufficient number of neurons.The proposed method adds and deletes neurons using the means error and frequency in use among neighboring neurons.