k-Propagated Self-organizing Maps

T. Yanagida and T. Miura (Japan)


Machine Learning, Artificial Neural Network (ANN), SOM(k), LVQ(k), classifiers


In this investigation, we discuss Self Organizing Maps (SOM) based classifiers with propagation on SOM map. To enhance explanation capability, we introduce neighbor learning and define k-propagated SOM, called SOM(k). We show experimental results to see how SOM(k) is useful. The main idea is that we extend SOM by propagating classification as well as parameter learning of SOM.

Important Links:

Go Back