|
wolffd@0
|
1 function bnet = mk_incinerator_bnet(ns)
|
|
wolffd@0
|
2 % MK_INCINERATOR_BNET The waste incinerator emissions example from Cowell et al p145
|
|
wolffd@0
|
3 % function bnet = mk_incinerator_bnet(ns)
|
|
wolffd@0
|
4 %
|
|
wolffd@0
|
5 % If ns is omitted, we use the scalars and binary nodes and the original params.
|
|
wolffd@0
|
6 % Otherwise, we use random params of the desired size.
|
|
wolffd@0
|
7 %
|
|
wolffd@0
|
8 % Lauritzen, "Propogation of Probabilities, Means and Variances in Mixed Graphical Association Models",
|
|
wolffd@0
|
9 % JASA 87(420): 1098--1108
|
|
wolffd@0
|
10 % This example is reprinted on p145 of "Probabilistic Networks and Expert Systems",
|
|
wolffd@0
|
11 % Cowell, Dawid, Lauritzen and Spiegelhalter, 1999, Springer.
|
|
wolffd@0
|
12 % For a picture, see http://www.cs.berkeley.edu/~murphyk/Bayes/usage.html#cg_model
|
|
wolffd@0
|
13
|
|
wolffd@0
|
14 % node numbers
|
|
wolffd@0
|
15 F = 1; W = 2; E = 3; B = 4; C = 5; D = 6; Min = 7; Mout = 8; L = 9;
|
|
wolffd@0
|
16 names = {'F', 'W', 'E', 'B', 'C', 'D', 'Min', 'Mout', 'L'};
|
|
wolffd@0
|
17 n = 9;
|
|
wolffd@0
|
18 dnodes = [F W B];
|
|
wolffd@0
|
19 cnodes = mysetdiff(1:n, dnodes);
|
|
wolffd@0
|
20
|
|
wolffd@0
|
21 % node sizes - all cts nodes are scalar, all discrete nodes are binary
|
|
wolffd@0
|
22 if nargin < 1
|
|
wolffd@0
|
23 ns = ones(1, n);
|
|
wolffd@0
|
24 ns(dnodes) = 2;
|
|
wolffd@0
|
25 rnd = 0;
|
|
wolffd@0
|
26 else
|
|
wolffd@0
|
27 rnd = 1;
|
|
wolffd@0
|
28 end
|
|
wolffd@0
|
29
|
|
wolffd@0
|
30 % topology (p 1099, fig 1)
|
|
wolffd@0
|
31 dag = zeros(n);
|
|
wolffd@0
|
32 dag(F,E)=1;
|
|
wolffd@0
|
33 dag(W,[E Min D]) = 1;
|
|
wolffd@0
|
34 dag(E,D)=1;
|
|
wolffd@0
|
35 dag(B,[C D])=1;
|
|
wolffd@0
|
36 dag(D,[L Mout])=1;
|
|
wolffd@0
|
37 dag(Min,Mout)=1;
|
|
wolffd@0
|
38
|
|
wolffd@0
|
39 % params (p 1102)
|
|
wolffd@0
|
40 bnet = mk_bnet(dag, ns, 'discrete', dnodes, 'names', names);
|
|
wolffd@0
|
41
|
|
wolffd@0
|
42 if rnd
|
|
wolffd@0
|
43 for i=dnodes(:)'
|
|
wolffd@0
|
44 bnet.CPD{i} = tabular_CPD(bnet, i);
|
|
wolffd@0
|
45 end
|
|
wolffd@0
|
46 for i=cnodes(:)'
|
|
wolffd@0
|
47 bnet.CPD{i} = gaussian_CPD(bnet, i);
|
|
wolffd@0
|
48 end
|
|
wolffd@0
|
49 else
|
|
wolffd@0
|
50 bnet.CPD{B} = tabular_CPD(bnet, B, 'CPT', [0.85 0.15]); % 1=stable, 2=unstable
|
|
wolffd@0
|
51 bnet.CPD{F} = tabular_CPD(bnet, F, 'CPT', [0.95 0.05]); % 1=intact, 2=defect
|
|
wolffd@0
|
52 bnet.CPD{W} = tabular_CPD(bnet, W, 'CPT', [2/7 5/7]); % 1=industrial, 2=household
|
|
wolffd@0
|
53 bnet.CPD{E} = gaussian_CPD(bnet, E, 'mean', [-3.9 -0.4 -3.2 -0.5], ...
|
|
wolffd@0
|
54 'cov', [0.00002 0.0001 0.00002 0.0001]);
|
|
wolffd@0
|
55 bnet.CPD{D} = gaussian_CPD(bnet, D, 'mean', [6.5 6.0 7.5 7.0], ...
|
|
wolffd@0
|
56 'cov', [0.03 0.04 0.1 0.1], 'weights', [1 1 1 1]);
|
|
wolffd@0
|
57 bnet.CPD{C} = gaussian_CPD(bnet, C, 'mean', [-2 -1], 'cov', [0.1 0.3]);
|
|
wolffd@0
|
58 bnet.CPD{L} = gaussian_CPD(bnet, L, 'mean', 3, 'cov', 0.25, 'weights', -0.5);
|
|
wolffd@0
|
59 bnet.CPD{Min} = gaussian_CPD(bnet, Min, 'mean', [0.5 -0.5], 'cov', [0.01 0.005]);
|
|
wolffd@0
|
60 bnet.CPD{Mout} = gaussian_CPD(bnet, Mout, 'mean', 0, 'cov', 0.002, 'weights', [1 1]);
|
|
wolffd@0
|
61 end
|
