-
Notifications
You must be signed in to change notification settings - Fork 17
/
hbayes.m
49 lines (46 loc) · 1.33 KB
/
hbayes.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
function [h, hdata] = hbayes(net, hdata)
%HBAYES Evaluate Hessian of Bayesian error function for network.
%
% Description
% H = HBAYES(NET, HDATA) takes a network data structure NET together
% the data contribution to the Hessian for a set of inputs and targets.
% It returns the regularised Hessian using any zero mean Gaussian
% priors on the weights defined in NET. In addition, if a MASK is
% defined in NET, then the entries in H that correspond to weights with
% a 0 in the mask are removed.
%
% [H, HDATA] = HBAYES(NET, HDATA) additionally returns the data
% component of the Hessian.
%
% See also
% GBAYES, GLMHESS, MLPHESS, RBFHESS
%
% Copyright (c) Ian T Nabney (1996-2001)
if (isfield(net, 'mask'))
% Extract relevant entries in Hessian
nmask_rows = size(find(net.mask), 1);
hdata = reshape(hdata(logical(net.mask*(net.mask'))), ...
nmask_rows, nmask_rows);
nwts = nmask_rows;
else
nwts = net.nwts;
end
if isfield(net, 'beta')
h = net.beta*hdata;
else
h = hdata;
end
if isfield(net, 'alpha')
if size(net.alpha) == [1 1]
h = h + net.alpha*eye(nwts);
else
if isfield(net, 'mask')
nindx_cols = size(net.index, 2);
index = reshape(net.index(logical(repmat(net.mask, ...
1, nindx_cols))), nmask_rows, nindx_cols);
else
index = net.index;
end
h = h + diag(index*net.alpha);
end
end