Package databionics.esom.train

Implementation of training algorithms.

See:
Description

Class Summary

Batch2SOM	batch version of SOM Training, without counting more hits, just update everything after whole epoche
BatchSOM	Implementation of the batch SOM training.
BestMatchHit	Represent the projection of one ore more data vectors on a best match.
Descriptives	Computes and stores important characteristics of the data
HybridBatchSOM	batch version, which goes online, if too many same hits are realized
KBatchSOM	version of batch training, updates map every k-th epoch
OnlineSOM	Implementation of the standard SOM training by pattern: The update of the best matching neuron and its neighbors is done right after the best match search for the curent pattern.
SlowBatchSOM	batch version of SOM Training, without counting more hits, just update everything after whole epoche
SOM	Abstract base class for a self organizing map training algorithm.
TrainAction	calls the dialog for training
TrainCallback	callback for trainig run dialog

Package databionics.esom.train Description

Implementation of training algorithms. All algorithms should inherit from SOM. The SOM class contains the basic algorithm broken down in elementary steps and methods for bestmatch search and neuron update. OnlineSOM and BatchSOM implement these steps in the corresponding order. Possible extensions of this package include algorithms for temporal SOM training or fast approximate training methods.