boxdist (Neural Network Toolbox)

Box distance function

d = boxdist(pos);

boxdist is a layer distance function that is used to find the distances between the layer's neurons, given their positions.

boxdist(pos) takes one argument,

pos - N x S matrix of neuron positions.

and returns the S x S matrix of distances.

boxdist is most commonly used in conjunction with layers whose topology function is gridtop.

Examples

Here we define a random matrix of positions for 10 neurons arranged in three-dimensional space and then find their distances.

pos = rand(3,10);
d = boxdist(pos)

You can create a standard network that uses boxdist as a distance function by calling newsom.

To change a network so that a layer's topology uses boxdist, set net.layers{i}.distanceFcn to 'boxdist'.

In either case, call sim to simulate the network with boxdist. See newsom for training and adaption examples.

The box distance D between two position vectors Pi and Pj from a set of S vectors is:

Dij = max(abs(Pi-Pj))