|
Daniel@0
|
1 <html>
|
|
Daniel@0
|
2 <head>
|
|
Daniel@0
|
3 <title>
|
|
Daniel@0
|
4 Netlab Reference Manual mdnfwd
|
|
Daniel@0
|
5 </title>
|
|
Daniel@0
|
6 </head>
|
|
Daniel@0
|
7 <body>
|
|
Daniel@0
|
8 <H1> mdnfwd
|
|
Daniel@0
|
9 </H1>
|
|
Daniel@0
|
10 <h2>
|
|
Daniel@0
|
11 Purpose
|
|
Daniel@0
|
12 </h2>
|
|
Daniel@0
|
13 Forward propagation through Mixture Density Network.
|
|
Daniel@0
|
14
|
|
Daniel@0
|
15 <p><h2>
|
|
Daniel@0
|
16 Synopsis
|
|
Daniel@0
|
17 </h2>
|
|
Daniel@0
|
18 <PRE>
|
|
Daniel@0
|
19 mixparams = mdnfwd(net, x)
|
|
Daniel@0
|
20 [mixparams, y, z] = mdnfwd(net, x)
|
|
Daniel@0
|
21 [mixparams, y, z, a] = mdnfwd(net, x)
|
|
Daniel@0
|
22 </PRE>
|
|
Daniel@0
|
23
|
|
Daniel@0
|
24
|
|
Daniel@0
|
25 <p><h2>
|
|
Daniel@0
|
26 Description
|
|
Daniel@0
|
27 </h2>
|
|
Daniel@0
|
28
|
|
Daniel@0
|
29 <CODE>mixparams = mdnfwd(net, x)</CODE> takes a mixture density network data
|
|
Daniel@0
|
30 structure <CODE>net</CODE> and a matrix <CODE>x</CODE> of input vectors, and forward
|
|
Daniel@0
|
31 propagates the inputs through the network to generate a structure
|
|
Daniel@0
|
32 <CODE>mixparams</CODE> which contains the parameters of several mixture models.
|
|
Daniel@0
|
33 Each row of <CODE>x</CODE> represents
|
|
Daniel@0
|
34 one input vector and the corresponding row of the matrices in <CODE>mixparams</CODE>
|
|
Daniel@0
|
35 represents the parameters of a mixture model for the conditional probability
|
|
Daniel@0
|
36 of target vectors given the input vector. This is not represented as an array
|
|
Daniel@0
|
37 of <CODE>gmm</CODE> structures to improve the efficiency of MDN training.
|
|
Daniel@0
|
38
|
|
Daniel@0
|
39 <p>The fields in <CODE>mixparams</CODE> are
|
|
Daniel@0
|
40 <PRE>
|
|
Daniel@0
|
41
|
|
Daniel@0
|
42 type = 'mdnmixes'
|
|
Daniel@0
|
43 ncentres = number of mixture components
|
|
Daniel@0
|
44 dimtarget = dimension of target space
|
|
Daniel@0
|
45 mixcoeffs = mixing coefficients
|
|
Daniel@0
|
46 centres = means of Gaussians: stored as one row per pattern
|
|
Daniel@0
|
47 covars = covariances of Gaussians
|
|
Daniel@0
|
48 nparams = number of parameters
|
|
Daniel@0
|
49 </PRE>
|
|
Daniel@0
|
50
|
|
Daniel@0
|
51
|
|
Daniel@0
|
52 <p><CODE>[mixparams, y, z] = mdnfwd(net, x)</CODE> also generates a matrix <CODE>y</CODE> of
|
|
Daniel@0
|
53 the outputs of the MLP and a matrix <CODE>z</CODE> of the hidden
|
|
Daniel@0
|
54 unit activations where each row corresponds to one pattern.
|
|
Daniel@0
|
55
|
|
Daniel@0
|
56 <p><CODE>[mixparams, y, z, a] = mlpfwd(net, x)</CODE> also returns a matrix <CODE>a</CODE>
|
|
Daniel@0
|
57 giving the summed inputs to each output unit, where each row
|
|
Daniel@0
|
58 corresponds to one pattern.
|
|
Daniel@0
|
59
|
|
Daniel@0
|
60 <p><h2>
|
|
Daniel@0
|
61 See Also
|
|
Daniel@0
|
62 </h2>
|
|
Daniel@0
|
63 <CODE><a href="mdn.htm">mdn</a></CODE>, <CODE><a href="mdn2gmm.htm">mdn2gmm</a></CODE>, <CODE><a href="mdnerr.htm">mdnerr</a></CODE>, <CODE><a href="mdngrad.htm">mdngrad</a></CODE>, <CODE><a href="mlpfwd.htm">mlpfwd</a></CODE><hr>
|
|
Daniel@0
|
64 <b>Pages:</b>
|
|
Daniel@0
|
65 <a href="index.htm">Index</a>
|
|
Daniel@0
|
66 <hr>
|
|
Daniel@0
|
67 <p>Copyright (c) Ian T Nabney (1996-9)
|
|
Daniel@0
|
68 <p>David J Evans (1998)
|
|
Daniel@0
|
69
|
|
Daniel@0
|
70 </body>
|
|
Daniel@0
|
71 </html> |
