Efficient Self-Organizing Map Learning Scheme Using Data Reduction Preprocessing

The traditional Self-Organizing Map usually considers the whole data set in one go, whereas the dominative representative data are not well utilized. The learning process is found to be rigid and time-consuming when one is dealing with large data sets. In this paper, we propose to apply density based data reduction method as preprocessing. The proposed method extracts representative data preliminarily for the SOM training, and it is found to be particularly useful in terms of reducing the overall computational time. The accuracy of the SOM map is gradually increased according to the relationship between the remaining data and the representatives. In this paper, comparative studies between our proposed method and the basic SOM are included. Simulation results on three data sets demonstrate that the newly proposed method is an efficient approach and it consistently outperforms the conventional training method.