|
Daniel@0
|
1 <html>
|
|
Daniel@0
|
2 <head>
|
|
Daniel@0
|
3 <title>
|
|
Daniel@0
|
4 Netlab Reference Manual gtm
|
|
Daniel@0
|
5 </title>
|
|
Daniel@0
|
6 </head>
|
|
Daniel@0
|
7 <body>
|
|
Daniel@0
|
8 <H1> gtm
|
|
Daniel@0
|
9 </H1>
|
|
Daniel@0
|
10 <h2>
|
|
Daniel@0
|
11 Purpose
|
|
Daniel@0
|
12 </h2>
|
|
Daniel@0
|
13 Create a Generative Topographic Map.
|
|
Daniel@0
|
14
|
|
Daniel@0
|
15 <p><h2>
|
|
Daniel@0
|
16 Synopsis
|
|
Daniel@0
|
17 </h2>
|
|
Daniel@0
|
18 <PRE>
|
|
Daniel@0
|
19 net = gtm(dimlatent, nlatent, dimdata, ncentres, rbfunc)
|
|
Daniel@0
|
20 net = gtm(dimlatent, nlatent, dimdata, ncentres, rbfunc, prior)
|
|
Daniel@0
|
21 </PRE>
|
|
Daniel@0
|
22
|
|
Daniel@0
|
23
|
|
Daniel@0
|
24 <p><h2>
|
|
Daniel@0
|
25 Description
|
|
Daniel@0
|
26 </h2>
|
|
Daniel@0
|
27
|
|
Daniel@0
|
28 <p><CODE>net = gtm(dimlatent, nlatent, dimdata, ncentres, rbfunc)</CODE>,
|
|
Daniel@0
|
29 takes the dimension of the latent space <CODE>dimlatent</CODE>, the
|
|
Daniel@0
|
30 number of data points sampled in the latent space <CODE>nlatent</CODE>, the
|
|
Daniel@0
|
31 dimension of the data space <CODE>dimdata</CODE>, the number of centres in the
|
|
Daniel@0
|
32 RBF model <CODE>ncentres</CODE>, the activation function for the RBF
|
|
Daniel@0
|
33 <CODE>rbfunc</CODE>
|
|
Daniel@0
|
34 and returns a data structure <CODE>net</CODE>. The parameters in the
|
|
Daniel@0
|
35 RBF and GMM sub-models are set by calls to the corresponding creation routines
|
|
Daniel@0
|
36 <CODE>rbf</CODE> and <CODE>gmm</CODE>.
|
|
Daniel@0
|
37
|
|
Daniel@0
|
38 <p>The fields in <CODE>net</CODE> are
|
|
Daniel@0
|
39 <PRE>
|
|
Daniel@0
|
40 type = 'gtm'
|
|
Daniel@0
|
41 nin = dimension of data space
|
|
Daniel@0
|
42 dimlatent = dimension of latent space
|
|
Daniel@0
|
43 rbfnet = RBF network data structure
|
|
Daniel@0
|
44 gmmnet = GMM data structure
|
|
Daniel@0
|
45 X = sample of latent points
|
|
Daniel@0
|
46 </PRE>
|
|
Daniel@0
|
47
|
|
Daniel@0
|
48
|
|
Daniel@0
|
49 <p><CODE>net = gtm(dimlatent, nlatent, dimdata, ncentres, rbfunc, prior)</CODE>,
|
|
Daniel@0
|
50 sets a Gaussian zero mean prior on the
|
|
Daniel@0
|
51 parameters of the RBF model. <CODE>prior</CODE> must be a scalar and represents
|
|
Daniel@0
|
52 the inverse variance of the prior distribution. This gives rise to
|
|
Daniel@0
|
53 a weight decay term in the error function.
|
|
Daniel@0
|
54
|
|
Daniel@0
|
55 <p><h2>
|
|
Daniel@0
|
56 See Also
|
|
Daniel@0
|
57 </h2>
|
|
Daniel@0
|
58 <CODE><a href="gtmfwd.htm">gtmfwd</a></CODE>, <CODE><a href="gtmpost.htm">gtmpost</a></CODE>, <CODE><a href="rbf.htm">rbf</a></CODE>, <CODE><a href="gmm.htm">gmm</a></CODE><hr>
|
|
Daniel@0
|
59 <b>Pages:</b>
|
|
Daniel@0
|
60 <a href="index.htm">Index</a>
|
|
Daniel@0
|
61 <hr>
|
|
Daniel@0
|
62 <p>Copyright (c) Ian T Nabney (1996-9)
|
|
Daniel@0
|
63
|
|
Daniel@0
|
64
|
|
Daniel@0
|
65 </body>
|
|
Daniel@0
|
66 </html> |
