Random Neural Net classifier
W = RNNC(A,N,S)
W is a feed-forward neural net with one hidden layer of N sigmoid neurons. The input layer rescales the input features to unit variance; the hidden layer has normally distributed weights and biases with zero mean and standard deviation S. The output layer is trained by the dataset A. Default N is number of objects * 0.2, but not more than 100.
If N and/or S is NaN they are optimised by REGOPTC.
Uses the Mathworks' Neural Network toolbox. Consequently it is not available under Octave.
1. W.F. Schmidt, M.A. Kraaijveld, and R.P.W. Duin, Feed forward neural networks with random weights, Proc. ICPR11, Volume II, 1992, 1-4.