Scientific paper ID 1374 : 2016/3
SELFORGANIZING ROAD MAP
The network is initialized with small random values or with the eigenvectors of the main components. Upon learning of the net sample values are passed to the input neurons and with the closest weight vector is located. The weight of thi sneuron closest to it on the network adapts to the input vector, such as the size of this adjustment depends on the distance between neurons in the network.
самоорганизираща се пътна мрежаselforganizing road mapValentina Hristova
Carmichael, G.,“MatchingSphericalPanoramas to PlanarPhotographs”, M.C.S. thesis, CarletonUniversity, Ottawa, Canada, 2009.