ART-based parallel learning of growing SOMs and its application to TSP
ICONIP'06 Proceedings of the 13 international conference on Neural Information Processing - Volume Part I
Parallel ant colony optimizers with local and global ants
IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
Parallel ant colony optimizer based on adaptive resonance theory maps
ICONIP'08 Proceedings of the 15th international conference on Advances in neuro-information processing - Volume Part I
Hi-index | 0.00 |
Collaboration of growing self-organizing maps (GSOM) and adaptive resonance theory maps (ART) is considered through traveling sales-person problems (TSP).The ART is used to parallelize the GSOM: it divides the input space of city positions into subspaces automatically. One GSOM is allocated to each subspace and grows following the input data. After all the GSOMs grow sufficiently they are connected and we obtain a tour. Basic experimental results suggest that we can find semi-optimal solution much faster than serial methods.