A non-greedy approach to tree-structured clustering

David Jonathan Miller, Kenneth Rose

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

We propose a new interdisciplinary approach for the hard optimization problem of tree-structured clustering, wherein the imposition of structural constraints on the solution drastically reduces the complexity of classifying data. The method, derived with analogy to statistical physics, performs a global optimization over the entire tree. Experimentation with non-trivial examples shows that our approach consistently outperforms greedy methods and avoids local minima that may trap them.

Original languageEnglish (US)
Pages (from-to)683-690
Number of pages8
JournalPattern Recognition Letters
Volume15
Issue number7
DOIs
StatePublished - Jan 1 1994

Fingerprint

Global optimization
Physics

All Science Journal Classification (ASJC) codes

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

@article{744fbae130f24827894f463235850c87,
title = "A non-greedy approach to tree-structured clustering",
abstract = "We propose a new interdisciplinary approach for the hard optimization problem of tree-structured clustering, wherein the imposition of structural constraints on the solution drastically reduces the complexity of classifying data. The method, derived with analogy to statistical physics, performs a global optimization over the entire tree. Experimentation with non-trivial examples shows that our approach consistently outperforms greedy methods and avoids local minima that may trap them.",
author = "Miller, {David Jonathan} and Kenneth Rose",
year = "1994",
month = "1",
day = "1",
doi = "10.1016/0167-8655(94)90072-8",
language = "English (US)",
volume = "15",
pages = "683--690",
journal = "Pattern Recognition Letters",
issn = "0167-8655",
publisher = "Elsevier",
number = "7",

}

A non-greedy approach to tree-structured clustering. / Miller, David Jonathan; Rose, Kenneth.

In: Pattern Recognition Letters, Vol. 15, No. 7, 01.01.1994, p. 683-690.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A non-greedy approach to tree-structured clustering

AU - Miller, David Jonathan

AU - Rose, Kenneth

PY - 1994/1/1

Y1 - 1994/1/1

N2 - We propose a new interdisciplinary approach for the hard optimization problem of tree-structured clustering, wherein the imposition of structural constraints on the solution drastically reduces the complexity of classifying data. The method, derived with analogy to statistical physics, performs a global optimization over the entire tree. Experimentation with non-trivial examples shows that our approach consistently outperforms greedy methods and avoids local minima that may trap them.

AB - We propose a new interdisciplinary approach for the hard optimization problem of tree-structured clustering, wherein the imposition of structural constraints on the solution drastically reduces the complexity of classifying data. The method, derived with analogy to statistical physics, performs a global optimization over the entire tree. Experimentation with non-trivial examples shows that our approach consistently outperforms greedy methods and avoids local minima that may trap them.

UR - http://www.scopus.com/inward/record.url?scp=0028467104&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028467104&partnerID=8YFLogxK

U2 - 10.1016/0167-8655(94)90072-8

DO - 10.1016/0167-8655(94)90072-8

M3 - Article

VL - 15

SP - 683

EP - 690

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

SN - 0167-8655

IS - 7

ER -