The 25th International Conference on Machine Learning (ICML 2008)

Spectral Clustering with Inconsistent Advice

author: Tom Coleman, The University of Melbourne

Description

Clustering with advice (often known as constrained clustering) has been a recent focus of the data mining community. Success has been achieved incorporating advice into the k-means framework, as well as spectral clustering. Although the theory community has explored inconsistent advice, it has not yet been incorporated into spectral clustering. Extending work of De Bie and Cristianini, we set out a framework for finding minimum normalized cuts, subject to inconsistent advice. Our results suggest that the framework will be successful in many situations.

You might be experiencing some problems with Your Video player.

*Slides*
0:00	Spectral Clustering with Inconsistent Advice
0:20	Outline
0:53	Clustering (1)
0:55	Clustering (2)
0:57	Clustering (3)
1:08	Clustering (4)
1:12	Clustering (5)
1:13	Clustering (6)
1:15	Input: Affinities (1)
1:19	Input: Affinities (2)
1:21	Input: Affinities (3)
1:23	Input: Affinities (4)
1:32	Input: Affinities (5)
1:37	Clustering Costs (1)
1:44	Clustering Costs (2)
1:47	Clustering Costs (3)
1:56	Clustering Costs (4)
2:08	Clustering Costs (5)
2:16	Clustering Costs (6)
2:46	Advice for Clustering Problems (1)
2:56	Advice for Clustering Problems (2)
3:02	Advice for Clustering Problems (3)
3:28	Advice for Clustering Problems (4)
3:31	Advice for Clustering Problems (5)
4:14	The 2-Correlation Clustering Problem (1)
4:22	The 2-Correlation Clustering Problem (2)
4:24	The 2-Correlation Clustering Problem (3)
4:30	The 2-Correlation Clustering Problem (4)
4:32	The 2-Correlation Clustering Problem (5)
4:35	The 2-Correlation Clustering Problem (6)
4:39	The 2-Correlation Clustering Problem (7)
4:46	Clustering with Inconsistent Advice (1)
4:51	Clustering with Inconsistent Advice (2)
5:00	Outline
5:23	Ratio Cut (1)
5:47	Ratio Cut (2)
7:38	Ratio Cut (3)
8:17	The Rayleigh-Ritz Theorem (1)
8:26	Ratio Cut (3)
8:30	The Rayleigh-Ritz Theorem (1)
8:43	The Rayleigh-Ritz Theorem (2)
8:51	The Rayleigh-Ritz Theorem (3)
9:01	The Rayleigh-Ritz Theorem (4)
9:22	Spectral Clustering Solution
9:55	Spectral Clustering with Subspace Constraints (1)
10:20	Spectral Clustering with Subspace Constraints (2)
10:28	Spectral Clustering with Subspace Constraints (3)
10:38	The ‘subspace trick’ [De Bie, Suykens, De Moor (2004)] (1)
10:54	The ‘subspace trick’ [De Bie, Suykens, De Moor (2004)] (2)
11:42	The ‘subspace trick’ [De Bie, Suykens, De Moor (2004)] (3)
11:59	An example
12:39	Method One (1)
13:07	Method One (2)
13:16	Method One (3)
13:23	Method One (4)
13:45	Method One (5)
14:35	Method Two (1)
15:00	Method Two (2)
15:27	Method Three
15:58	Spectral relaxation of 2CC (1)
16:24	Spectral relaxation of 2CC (2)
16:26	Spectral relaxation of 2CC (3)
16:58	Spectral relaxation of 2CC (4)
17:00	A natural family of subspaces (1)
17:20	A natural family of subspaces (2)
17:23	Summary of solution method
17:51	Outline
17:54	Datasets (1)
17:58	Datasets (2)
18:28	Datasets (3)
18:44	Datasets (4)
19:07	Datasets (5)
19:31	Heart Disease, Dense Advice, p=0.75
20:59	Heart Disease, Complete Advice
21:46	Conclusions (1)
21:58	Conclusions (2)
22:13	Conclusions (3)
22:15	Conclusions (4)
22:33	Conclusions (5)
22:42	Conclusions (6)
22:47	Conclusions (7)

Lecture rating

People found this lecture:
Worth seeing
because it is:
Valuable and informative
Well presented
Easily understandable
Acceptably recorded
You need to login to cast your vote.

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.