This command trains a backpropagation network using the Matlab style of training algorithm. The method is based on a global adaptive learning rate parameter and on each momentum term is a weight-update procedure. By default one neuron for each input is used.
Example (ex5.9): Train a three-layered (input + hidden + output layer) network with a sine function using a sigmoid activation function in neurons. After training save the network output.
NDA> load sin.dat NDA> select sinx -f sin.x NDA> select siny -f sin.y NDA> mlbp -di sinx -do siny -net 3 1 15 1 -types t t t -em 100 -nout wei -ef virhe -bs 0.1 -ac 1.1 -mup 1.1 -mdn 0.7 -mom 0.95 NDA> fbp -di sinx -do out -win wei NDA> select output -f sin.x out.0 NDA> save output
