A learning model of the feature-detecting cells for unsupervised pattern classification

A modified neural network unsupervised learning scheme by the feature-detecting cell is proposed. We improve the performance in learning categories by adding a modulation system and a competing system to the conventional feature-detecting cell model. With the aid of the modulation system and the competing system, the cluster prototype which is closest to the input pattern will win the competitions and a winner dominated learning will be controlled by the properly assigned bias values. The proposed model has the following features: it guarantees the corresponding feature-detecting cell of each input pattern to be formed regardless of the initial weights and the duplication (of the feature-detecting cells formation for each sampled input pattern) can be reduced.