Kernels for Link Prediction with Latent Feature Models thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Kernels for Link Prediction with Latent Feature Models

Published on Oct 03, 20113293 Views

Predicting new links in a network is a problem of interest in many application domains. Most of the prediction methods utilize information on the network’s entities such as nodes to build a model of l

Related categories

Chapter list

Kernels for Link Prediction with Latent Feature Models00:00
Outline00:16
Link Prediction on Networks00:38
Assumption of Link Prediction (1)01:38
Assumption of Link Prediction (2)02:12
Network Structures02:22
Objective02:53
Outline: Latent Feature Models03:57
Biological Motivation (1)04:04
Biological Motivation (2)04:54
Latent Feature Models of Links05:26
Inferring the Models06:21
Outline: Link Kernels07:14
Latent Feature Models in Kernels (1)07:22
Latent Feature Models in Kernels (2)07:54
Latent Feature Models in Kernels (3)08:09
Node Kernels (1)08:26
Node Kernels (2)08:53
Kernels on Links09:10
Demonstration09:32
Properties (Kn similar to ideal kernels)11:07
Properties (Kn similar to ideal kernels) cont.11:43
Outline: Experiments13:01
Experiments and Data13:11
On the Latent Feature Model Assumpsion13:42
Execution Time of Our Method vs. IBP14:38
Link Prediction Performance on yeast PPI networks15:39
Link Prediction Performance on fruit y PPI networks16:58
Conclusion (1)17:28
Conclusion (2)17:53