Selective Sampling on Graphs for Classification thumbnail
slide-image
Play
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Selective Sampling on Graphs for Classification

Published on Sep 27, 20134844 Views

Selective sampling is an active variant of online learning in which the learner is allowed to adaptively query the label of an observed example. The goal of selective sampling is to achieve a good

Related categories

Chapter list

Selective Sampling on Graphs for Classification00:00
Motivation01:14
Learning on Graphs01:16
Online Learning03:00
Active Learning04:09
Our Goal04:56
Outline - 105:50
Problem Setting of Batch Learning on Graphs05:53
Learning with Local and Global Consistency (LLGC)06:56
Ridge Regression Formulation of LLGC08:26
Outline - 208:48
Problem Setting of Online Learning on Graphs08:52
Algorithm 1 (OLLGC)09:19
Regret Bound of OLLGC10:37
Mistake Bound of OLLGC10:59
Outline - 311:39
Problem Setting of Selective Sampling on Graphs11:42
Key Idea12:22
Algorithm 2 (SSLGC)13:08
Regret Bound of SSLGC13:51
Label Complexity of SSLGC13:59
Outline - 414:45
Datasets14:48
Evaluation Measures15:30
Compared Methods16:00
A comparison of online learning and selective sampling algorithms on graphs - 116:22
A comparison of online learning and selective sampling algorithms on graphs - 217:16
Study on the Impact of ϰ17:49
Outline - 518:17
Summary18:18