annotate toolboxes/FullBNT-1.0.7/nethelp3.3/mlpbkp.htm @ 0:e9a9cd732c1e tip

first hg version after svn
author wolffd
date Tue, 10 Feb 2015 15:05:51 +0000
parents
children
rev   line source
wolffd@0 1 <html>
wolffd@0 2 <head>
wolffd@0 3 <title>
wolffd@0 4 Netlab Reference Manual mlpbkp
wolffd@0 5 </title>
wolffd@0 6 </head>
wolffd@0 7 <body>
wolffd@0 8 <H1> mlpbkp
wolffd@0 9 </H1>
wolffd@0 10 <h2>
wolffd@0 11 Purpose
wolffd@0 12 </h2>
wolffd@0 13 Backpropagate gradient of error function for 2-layer network.
wolffd@0 14
wolffd@0 15 <p><h2>
wolffd@0 16 Synopsis
wolffd@0 17 </h2>
wolffd@0 18 <PRE>
wolffd@0 19 g = mlpbkp(net, x, z, deltas)</PRE>
wolffd@0 20
wolffd@0 21
wolffd@0 22 <p><h2>
wolffd@0 23 Description
wolffd@0 24 </h2>
wolffd@0 25 <CODE>g = mlpbkp(net, x, z, deltas)</CODE> takes a network data structure
wolffd@0 26 <CODE>net</CODE> together with a matrix <CODE>x</CODE> of input vectors, a matrix
wolffd@0 27 <CODE>z</CODE> of hidden unit activations, and a matrix <CODE>deltas</CODE> of the
wolffd@0 28 gradient of the error function with respect to the values of the
wolffd@0 29 output units (i.e. the summed inputs to the output units, before the
wolffd@0 30 activation function is applied). The return value is the gradient
wolffd@0 31 <CODE>g</CODE> of the error function with respect to the network
wolffd@0 32 weights. Each row of <CODE>x</CODE> corresponds to one input vector.
wolffd@0 33
wolffd@0 34 <p>This function is provided so that the common backpropagation algorithm
wolffd@0 35 can be used by multi-layer perceptron network models to compute
wolffd@0 36 gradients for mixture density networks as well as standard error
wolffd@0 37 functions.
wolffd@0 38
wolffd@0 39 <p><h2>
wolffd@0 40 See Also
wolffd@0 41 </h2>
wolffd@0 42 <CODE><a href="mlp.htm">mlp</a></CODE>, <CODE><a href="mlpgrad.htm">mlpgrad</a></CODE>, <CODE><a href="mlpderiv.htm">mlpderiv</a></CODE>, <CODE><a href="mdngrad.htm">mdngrad</a></CODE><hr>
wolffd@0 43 <b>Pages:</b>
wolffd@0 44 <a href="index.htm">Index</a>
wolffd@0 45 <hr>
wolffd@0 46 <p>Copyright (c) Ian T Nabney (1996-9)
wolffd@0 47
wolffd@0 48
wolffd@0 49 </body>
wolffd@0 50 </html>