randnr (Neural Network Toolbox)

Normalized row weight initialization function

W = randnr(S,PR)
W = randnr(S,R)

randnr is a weight initialization function.

randnr(S,P) takes two inputs,

S - Number of rows (neurons).

PR - R x 2 matrix of input value ranges = [Pmin Pmax].

and returns an S x R random matrix with normalized rows.

Can also be called as randnr(S,R).

Examples

A matrix of three normalized four-element rows is generated:

M = randnr(3,4)
M =
    0.9713    0.0800   -0.1838   -0.1282
    0.8228    0.0338    0.1797    0.5381
   -0.3042   -0.5725    0.5436    0.5331