ABSTRACT
This chapter explains the hybrid self-organising map (SOM)–neural gas (NG) algorithm and its applications. It introduces classic algorithms and examines an understanding of the different features. The chapter presents the hybrid algorithm along with a step-by-step explanation of how a self-organising map algorithm can be created. It provides some examples and explores a how-to way for a deeper understanding of the uses. NG proposes a new way of calculating the influence, instead of using distance measures. NG is a capable algorithm and other versions were developed to achieve different kinds of results besides quantisation. SOM is a widely used algorithm for data processing and is very useful for a number of applications. The hybrid SOM-NG algorithm is a combination of SOM and NG, as the name clearly states. The hybrid SOM–NG algorithm is good for quantisation, but it can be used for many other applications, alone or combined with other techniques.
