Multilayer backpropagation with NETLAB toolbox

Multilayer backpropagation with NETLAB toolbox

I am trying to use the NETLAB toolbox to train a 3-layer
(input,hidden,output) feed-forward backpropagation Neural Network.
Unfortunately I do not have too much freedom in terms of network
architecture I can work with.
I notice NETLAB has the following functions that I need:
mlp,mlpbkp,mlpfwd,mlpgrad. I am not sure in what order I need to call the
above functions to train the network. The help manual is not of too much
help either.
If any of you have used the NETLAB toolbox, kindly let me know. Also, if
you know of other free toolboxes I can use in lieu of NETLAB, kindly, let
me know.
Thanks!