For the Self Organizing Map, what is the default neighborhood function?

When it comes to the self organizing map, for example the way it is done using the selforgmap function, what is the default neighborhood radius function and learning restraint? I had been digging through the functions that comprise the self organizing map function and was not sure that I could tell where they are. When there is a target weight vector W(q) for iteration q, we can start with the weight updating function W(q) = W(q-1) + N(q,v,s)*a(s)*(D(t) - W(q-1)) where t is the index of a given input weight vector, D(t) are the input vector's weights, v is the index of a map vector and s is the index of the BMU. Naturally, N(q,v,s) is the neighborhood function and a(s) is the learning restraint.

For the matlab function selforgmap, what is the default function used for N(q,v,s) and a(s)? I was looking to see where these functions are defined in the selforgmap code but was not sure. What are the default functions used?

Answer 1

[gunjan]

https://www.mathworks.com/help/nnet/ug/cluster-with-self-organizing-map-neural-network.html#bss4b_l-3

check this.. it says that the weights are updated : W(k+1) = W(k) + alpha * [ (dist) ] rule.. where dist = distance between current neuron that you are updating and the winning neuron. k = iteration number alpha = learning rate