
en
0.25
0.5
0.75
1.25
1.5
1.75
2
No voodoo here! Learning discrete graphical models via inverse covariance estimation
Published on 2013-01-166445 Views
We investigate the relationship between the support of the inverses of generalized covariance matrices and the structure of a discrete graphical model. We show that for certain graph structures, the s
Related categories
Presentation
No voodoo here! Learning discrete graphical models via inverse covariance estimation00:00
Introduction: Graphical models (1)00:22
Introduction: Graphical models (2)00:45
Introduction: Graphical models (3)01:02
Introduction: Graphical models (4)01:14
Introduction: Graphical models (5)01:38
Structure learning for Gaussians (1)02:01
Structure learning for Gaussians (2)02:41
Non-Gaussian distributions (1)03:49
Non-Gaussian distributions (2)04:02
Non-Gaussian distributions (3)04:28
A curious example (1)04:38
A curious example (2)04:55
A curious example (3)05:04
A curious example (4)05:10
A curious example (5)05:26
A curious example (6)06:15
A curious example (7)06:23
A curious example (8)06:53
Some notation (1)07:56
Some notation (2)08:13
Some notation (3)08:44
Some notation (4)09:05
Augmenting covariance matrix (1)09:22
Augmenting covariance matrix (2)09:32
Augmenting covariance matrix (3)09:45
Augmenting covariance matrix (4)10:08
Augmenting covariance matrix (5)10:45
Example: Binary Ising model (1)10:50
Example: Binary Ising model (2)11:07
Consequences for trees (1)12:03
Consequences for trees (2)12:13
Consequences for trees (3)12:26
Structure learning (1)12:44
Structure learning (2)13:04
Structure learning (3)13:43
Graphical Lasso (1)13:55
Graphical Lasso (2)14:30
Graphical Lasso (3)14:58
Simulation study15:23
Inference methods for non-trees (1)15:51
Inference methods for non-trees (2)16:20
Inference methods for non-trees (3)16:24
Summary (1)16:34
Summary (2)16:51
Summary (3)16:56
Summary (4)17:07