Software Tools for SOM

In order to obtain a “hands-on” experience of the methods, many people like to program neural network algorithms, such as the SOM, themselves. This is in general a good practice; however, many important factors should be taken into account. Neural-network methods in general, and the SOM process in particular do not produce such unique results as some deterministic mathematical algorithms, e.g., the Fast Fourier Transform. Adaptive processes may proceed in unexpected ways depending on the parameters and training sequences selected. At least in the beginning, the learning process may seem to fluctuate in an unpredictable fashion. Therefore special precautions must be taken.

[1]  Juha Vesanto,et al.  SOM-based data visualization methods , 1999, Intell. Data Anal..