Geometric clustering using the Information Bottleneck method

Susanne Still; William Bialek; Léon Bottou

Geometric clustering using the Information Bottleneck method

Susanne Still, William Bialek, Léon Bottou

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We argue that K-means and deterministic annealing algorithms for geometric clustering can be derived from the more general Information Bottleneck approach. If we cluster the identities of data points to preserve information about their location, the set of optimal solutions is massively degenerate. But if we treat the equations that define the optimal solution as an iterative algorithm, then a set of "smooth" initial conditions selects solutions with the desired geometrical properties. In addition to conceptual unification, we argue that this approach can be more efficient and robust than classic algorithms.

Original language	English (US)
Title of host publication	Advances in Neural Information Processing Systems 16 - Proceedings of the 2003 Conference, NIPS 2003
Publisher	Neural information processing systems foundation
ISBN (Print)	0262201526, 9780262201520
State	Published - 2004
Event	17th Annual Conference on Neural Information Processing Systems, NIPS 2003 - Vancouver, BC, Canada Duration: Dec 8 2003 → Dec 13 2003

Publication series

Name	Advances in Neural Information Processing Systems
ISSN (Print)	1049-5258

Other

Other	17th Annual Conference on Neural Information Processing Systems, NIPS 2003
Country/Territory	Canada
City	Vancouver, BC
Period	12/8/03 → 12/13/03

All Science Journal Classification (ASJC) codes

Computer Networks and Communications
Information Systems
Signal Processing

Cite this

@inproceedings{138a471578f04f80a8640097b6770364,

title = "Geometric clustering using the Information Bottleneck method",

abstract = "We argue that K-means and deterministic annealing algorithms for geometric clustering can be derived from the more general Information Bottleneck approach. If we cluster the identities of data points to preserve information about their location, the set of optimal solutions is massively degenerate. But if we treat the equations that define the optimal solution as an iterative algorithm, then a set of {"}smooth{"} initial conditions selects solutions with the desired geometrical properties. In addition to conceptual unification, we argue that this approach can be more efficient and robust than classic algorithms.",

author = "Susanne Still and William Bialek and L{\'e}on Bottou",

year = "2004",

language = "English (US)",

isbn = "0262201526",

series = "Advances in Neural Information Processing Systems",

publisher = "Neural information processing systems foundation",

booktitle = "Advances in Neural Information Processing Systems 16 - Proceedings of the 2003 Conference, NIPS 2003",

note = "17th Annual Conference on Neural Information Processing Systems, NIPS 2003 ; Conference date: 08-12-2003 Through 13-12-2003",

}

Still, S, Bialek, W & Bottou, L 2004, Geometric clustering using the Information Bottleneck method. in Advances in Neural Information Processing Systems 16 - Proceedings of the 2003 Conference, NIPS 2003. Advances in Neural Information Processing Systems, Neural information processing systems foundation, 17th Annual Conference on Neural Information Processing Systems, NIPS 2003, Vancouver, BC, Canada, 12/8/03.

Geometric clustering using the Information Bottleneck method. / Still, Susanne; Bialek, William; Bottou, Léon.
Advances in Neural Information Processing Systems 16 - Proceedings of the 2003 Conference, NIPS 2003. Neural information processing systems foundation, 2004. (Advances in Neural Information Processing Systems).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Geometric clustering using the Information Bottleneck method

AU - Still, Susanne

AU - Bialek, William

AU - Bottou, Léon

PY - 2004

Y1 - 2004

N2 - We argue that K-means and deterministic annealing algorithms for geometric clustering can be derived from the more general Information Bottleneck approach. If we cluster the identities of data points to preserve information about their location, the set of optimal solutions is massively degenerate. But if we treat the equations that define the optimal solution as an iterative algorithm, then a set of "smooth" initial conditions selects solutions with the desired geometrical properties. In addition to conceptual unification, we argue that this approach can be more efficient and robust than classic algorithms.

AB - We argue that K-means and deterministic annealing algorithms for geometric clustering can be derived from the more general Information Bottleneck approach. If we cluster the identities of data points to preserve information about their location, the set of optimal solutions is massively degenerate. But if we treat the equations that define the optimal solution as an iterative algorithm, then a set of "smooth" initial conditions selects solutions with the desired geometrical properties. In addition to conceptual unification, we argue that this approach can be more efficient and robust than classic algorithms.

UR - http://www.scopus.com/inward/record.url?scp=84898998530&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84898998530&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84898998530

SN - 0262201526

SN - 9780262201520

T3 - Advances in Neural Information Processing Systems

BT - Advances in Neural Information Processing Systems 16 - Proceedings of the 2003 Conference, NIPS 2003

PB - Neural information processing systems foundation

T2 - 17th Annual Conference on Neural Information Processing Systems, NIPS 2003

Y2 - 8 December 2003 through 13 December 2003

ER -

Geometric clustering using the Information Bottleneck method

Abstract

Publication series

Other

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this