Rival Penalized Self-Organizing Map

L.-T. Law and Y.-M. Cheung (PRC)


Self-Orgainzing Map, Rival Penalization Controlled Com petitive Learning, Rival Penalized Self-Organizing Map.


Kohonen's Self-Organizing Map (SOM) is one of the most commonly used competitive learning algorithms that pro vide a topological mapping from the input space to the out put space. In the conventional SOM, it needs to choose an appropriate learning rate as well as a monotonically de creasing function that lowers the learning rate with time to ensure the convergence of the map. Otherwise, its perfor mance may seriously deteriorate. In this paper, we there fore propose a novel Rival Penalized Self-Organizing Map (RPSOM) learning algorithm, which dynamically penal izes a set of rivals towards driving far away from the in put data set during the learning. Compared to the existing methods, this new one need not select the monotonically decreasing function of the learning rate, but still gives a ro bust result. The experiments have shown its outstanding performance in comparison with the existing algorithms.

Important Links:

Go Back