en-de
en-es
en-fr
en-sl
en
en-zh
0.25
0.5
0.75
1.25
1.5
1.75
2
An Information Theoretic Approach to Learning Generative Graph Prototypes
Published on Oct 17, 20113412 Views
We present a method for constructing a generative model for sets of graphs by adopting a minimum description length approach. The method is posed in terms of learning a generative supergraph model fro
Related categories
Chapter list
An Information Theoretic Approach to Learning Generative Graph Prototypes00:00
Graph representations00:41
Graph representations from images01:34
Learning with graphs02:46
….. is difficult because03:59
Generative models04:29
Deep learning05:41
Aim06:54
Prior work07:12
Structural learning07:15
Description length07:40
Similarities/differences08:06
Coding scheme08:50
Method09:27
Model overview11:04
Learn supergraph using MDL12:04
Probabilistic framework12:37
Observation model12:43
Data code-length13:55
Information theory14:14
Von-Neumann entropy15:00
Approximation15:54
Computing traces16:37
Simplified entropy17:06
Overall code-length17:20
EM – code-length criterion17:34
Expectation + Maximization18:05
Experiments18:39
Experiments - validation19:00
Experiments - classification task20:14
Experiments - graph embedding (1)21:10
Experiments - graph embedding (2)21:49
Experiments - generate new samples22:14
Conclusion23:02