en-de
en-es
en-fr
en-pt
en-sl
en
en-zh
0.25
0.5
0.75
1.25
1.5
1.75
2
On Max-Margin Markov Networks in Hierarchical Document Classification
Published on Feb 25, 20074965 Views
Related categories
Chapter list
On Max-Margin Markov Networks in Hierarchical Document Classification00:01
Hierarchical Multilabel Classifcation: union of partial paths model00:36
Frequently used learning strategies for hierarchies01:53
The classification model03:43
Feature vectors04:45
Loss functions for hierarchies06:51
Max-margin Structured output learning10:09
Optimization problem11:25
Marginalized problem pt 112:36
Marginalized problem pt 213:47
Decomposing the model15:18
Conditional Gradient method16:14
Conditional Gradient Ascent pt 116:44
Conditional Gradient Ascent pt 216:54
Conditional Gradient Ascent pt 317:08
Conditional Gradient Ascent pt 417:16
Conditional Gradient Ascent pt 517:17
Conditional Gradient Ascent pt 617:18
Using inference to find update directions17:32
Experiments18:27
Optimization efficiency19:22
Prediction accuracy: Levelwise F120:05
Scalability?21:59
Conclusions23:03