Download presentation
Presentation is loading. Please wait.
1
1 White Parts from: Technical overview for machine-learning researcher – slides from UAI 1999 tutorialTechnical overview for machine-learning researcher – slides from UAI 1999 tutorial Part II
2
2
3
3
4
4 = C t,h Example: for (ht + htthh), we get p(d|m) = 3!2!/6!
5
5
6
6 Numerical example for the network X 1 X 2 Imaginary sample sizes denoted N’ ijk Data: (true, true) and (true, false)
7
7
8
8 Used so far Desired
9
9 How do we assign structure and parameter priors ? Structure priors: Uniform, partial order (allowed/prohibited edges), proportional to similarity to some a priori network.
10
10 BDe K2K2
11
11
12
12
13
13
14
14
15
15
16
16
17
17
18
18 Example: Suppose the hyper distribution for (X 1,X 2 ) is Dir( a 00, a 01,a 10, a 11 ). So how to generate parameter priors?
19
19 Example: Suppose the hyper distribution for (X 1,X 2 ) is Dir( a 00, a 01,a 10, a 11 ) This determines a Dirichlet distribution for the parameters of both directed models.
20
20
21
21 Summary: Suppose the parameters for (X 1,X 2 ) are distributed Dir( a 00, a 01,a 10, a 11 ). Then, parameters for X 1 are distributed Dir(a 00 +a 01,a 10 +a 11 ). Similarly, parameters for X 2 are distributed Dir(a 00 +a 10,a 01 +a 11 ).
22
22 BDe score:
23
23
24
24
25
25
26
26 n Example: f(x+y) = f(x) f(y) n Solution: (ln f )`(x+y) = (ln f )`(x) n and so: (ln f )`(x) = constant n Hence: (ln f )(x) = linear function n hence: f(x) = c e ax n Assumptions: Positive everywhere, Differentiable Functional Equations Example
27
27 The bivariate discrete case
28
28 The bivariate discrete case
29
29 The bivariate discrete case
30
30 The bivariate discrete case
31
31
32
32
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.