Representation of information using Kohonen's SOM (Self-Organizing Maps)

In this paper is presented a demonstration of Kohonen's self-organizing maps, also known as SOM. Likewise is prepared a study of the functioning of Kohonen's maps in one and two dimensions and the most important characteristics of this type of network that works in similar way that the human brain. Finally, this paper details the characteristics necessaries for the network's training and how is possible use the results of the neural networks to discover the characteristics of the information input for instance, how is your distribution, the densityand shape.