How can I use SOM (self-organizing map) for knowing deviations?
I want (may be wrong also) to use SOM for knowing deviation in a pattern.
The idea is I want to train SOM with some examples (un-supervised), and
I want to use new examples for prediction (whether pattern is same as past, i.e matching with already trained pattern or not).
Is it possible through RapidMiner?, I am unable to find the way.
Please help me.