Output details

Snap–drift neural network for self-organisation and sequence learning

<24>This paper represents the first time the snap-drift neural network has been applied to self-organising maps and to sequence learning. The combination of snap learning with learning vector quantisation yields demonstrably greater separation of clusters than traditional self-organising maps. Snap-drift is also a novel method for sequence learning in the context of the well known Simple Recurrent Network architecture and the results are very promising, both in terms of speed of learning (as measured by the required number of computations) and in terms of the effectiveness of the sequence learning and generalisation.