Selbstorganisierende Karte. Als Selbstorganisierende Karten, Kohonenkarten oder Kohonennetze (nach Teuvo Kohonen; englisch self-organizing map, SOM bzw. self-organizing feature map, SOFM) bezeichnet man eine Art von künstlichen neuronalen Netzen. Sie sind als unüberwachtes Lernverfahren ein leistungsfähiges Werkzeug des Data-Mining. Ihr Funktionsprinzip beruht auf der biologischen Erkenntnis, dass viele Strukturen im Gehirn eine lineare oder planare Topologie aufweisen. Die Signale des Eingangsraums, z. B. visuelle Reize, sind jedoch multidimensional. Es stellt sich also die Frage, wie diese multidimensionalen Eindrücke durch planare Strukturen verarbeitet werden. Biologische Untersuchungen zeigen, dass die Eingangssignale so abgebildet werden, dass ähnliche Reize nahe beieinander liegen.
Der Phasenraum der angelegten Reize wird also kartiert. Wird nun ein Signal an diese Karte herangeführt, so werden nur diejenigen Gebiete der Karte erregt, die dem Signal ähnlich sind. Laterale Umfeldhemmung[Bearbeiten] SOM tutorial part 1. Kohonen's Self Organizing Feature Maps Introductory Note This tutorial is the first of two related to self organising feature maps. Initially, this was just going to be one big comprehensive tutorial, but work demands and other time constraints have forced me to divide it into two. Nevertheless, part one should provide you with a pretty good introduction. Certainly more than enough to whet your appetite anyway! I will appreciate any feedback you are willing to give - good or bad. My ears are always open for praise, constructive criticism or suggestions for future tutorials.
Overview Kohonen Self Organising Feature Maps, or SOMs as I shall be referring to them from now on, are fascinating beasts. A common example used to help teach the principals behind SOMs is the mapping of colours from their three dimensional components - red, green and blue, into two dimensions. Figure 1 Screenshot of the demo program (left) and the colours it has classified (right).
Network Architecture Figure 2 Figure 3. SOM.pdf (application/pdf-Objekt) Self Organizing Map AI for Pictures. Introduction this article is about creating an app to cluster and search for related pictures. i got the basic idea from a Longhorn demo in which they showed similar functionality. in the demo, they selected an image of the sunset, and the program was able to search the other images on the hard drive and return similar images. there are other photo library applications that offer similar functionality. honestly ... i thought that was pretty cool, and wanted to have some idea how they might be doing that. internally, i do not know how they actually operate ... but this article will show one possibility. also writing this article to continue my AI training Kohonen SOM luckily there is a type of NN that works with unsupervised training. i'm guessing that it is the 2nd or 3rd most popular type of NN?
Anyways, that is my current understanding; here are some other articles i recommend 1) Grid Layout 2) Color Grouping 3) Blog Community (OUCH!) 4) Picture Similarity. Amazedsaint's .net journal.