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
50
51
|
function [h, hdata] = mlphess(net, x, t, hdata)
%MLPHESS Evaluate the Hessian matrix for a multi-layer perceptron network.
%
% Description
% H = MLPHESS(NET, X, T) takes an MLP network data structure NET, a
% matrix X of input values, and a matrix T of target values and returns
% the full Hessian matrix H corresponding to the second derivatives of
% the negative log posterior distribution, evaluated for the current
% weight and bias values as defined by NET.
%
% [H, HDATA] = MLPHESS(NET, X, T) returns both the Hessian matrix H and
% the contribution HDATA arising from the data dependent term in the
% Hessian.
%
% H = MLPHESS(NET, X, T, HDATA) takes a network data structure NET, a
% matrix X of input values, and a matrix T of target values, together
% with the contribution HDATA arising from the data dependent term in
% the Hessian, and returns the full Hessian matrix H corresponding to
% the second derivatives of the negative log posterior distribution.
% This version saves computation time if HDATA has already been
% evaluated for the current weight and bias values.
%
% See also
% MLP, HESSCHEK, MLPHDOTV, EVIDENCE
%
% Copyright (c) Ian T Nabney (1996-2001)
% Check arguments for consistency
errstring = consist(net, 'mlp', x, t);
if ~isempty(errstring);
error(errstring);
end
if nargin == 3
% Data term in Hessian needs to be computed
hdata = datahess(net, x, t);
end
[h, hdata] = hbayes(net, hdata);
% Sub-function to compute data part of Hessian
function hdata = datahess(net, x, t)
hdata = zeros(net.nwts, net.nwts);
for v = eye(net.nwts);
hdata(find(v),:) = mlphdotv(net, x, t, v);
end
return
|
