Daniel@0: Daniel@0:
Daniel@0:Daniel@0: demtrainDaniel@0: Daniel@0: Daniel@0:
demtrain
brings up a simple GUI to show the training of
Daniel@0: an MLP network on classification and regression problems. The user
Daniel@0: should load in a dataset (which should be in Netlab format: see
Daniel@0: datread
), select the output activation function, the
Daniel@0: number of cycles and hidden units and then
Daniel@0: train the network. The scaled conjugate gradient algorithm is used.
Daniel@0: A graph shows the evolution of the error: the value is shown
Daniel@0: max(ceil(iterations / 50), 5)
cycles.
Daniel@0:
Daniel@0: Once the network is trained, it is saved to the file mlptrain.net
.
Daniel@0: The results can then be viewed as a confusion matrix (for classification
Daniel@0: problems) or a plot of output versus target (for regression problems).
Daniel@0:
Daniel@0:
confmat
, datread
, mlp
, netopt
, scg
Copyright (c) Ian T Nabney (1996-9) Daniel@0: Daniel@0: Daniel@0: Daniel@0: