http://dsp.vscht.cz/konference_matlab/MATLAB08/prispevky/025_dostal.pdf

https://www.mathworks.com/help/nnet/gs/cluster-data-with-a-self-organizing-map.html

http://lab.fs.uni-lj.si/lasin/wp/IMIT_files/neural/NN-examples.pdf

https://www.academia.edu/34938587/Neural_Network_Toolbox_Users_Guide

https://archive.org/stream/NeuralNetworksUsingMatlab.ClusterAnalysisAndClassification/Neural%20Networks%20Using%20Matlab.%20Cluster%20Analysis%20and%20Classification_djvu.txt

The cluster analysis represents a group of methods whose aim is to classify the investigated objects into clusters. The founders of cluster analysis were Tryon, Ward and James. There have been suggested many new algorithms recently. Some methods represent a modification of classical methods of cluster analysis; other ones use advanced methods such as neural networks, e.g. represented by

Kohonen self-organizing maps, or genetic algorithms.

For more information please study the following:

* https://www.mathworks.com/help/deeplearning/ug/cluster-with-self-organizing-map-neural-network.html

* CLUSTER ANALYSIS AND NEURAL NETWORK.

In neural networks, the most commonly used network for clustering is the SOM. Mainly there are three measures of SOM performance; quantization error, topographic error, and distortion measure.

I recommend you to read Chapter 22 (Practical Training Issues) of the book "NN Design", by Martin T. Hagan, et al.