LearnMet

Learning domain-specific distance metrics for plots of scientific functions

Aparna Varde, Elke Rundensteiner, Carolina Ruiz, Mohammed Maniruzzaman, Richard Sisson

Research output: Contribution to journalArticleResearchpeer-review

2 Citations (Scopus)

Abstract

Scientific experimental results are often depicted as plots of functions to aid their visual analysis and comparison. In computationally comparing these plots using techniques such as similarity search and clustering, the notion of similarity is typically distance. However, it is seldom known which distance metric(s) best preserve(s) semantics in the respective domain. It is thus desirable to learn such domain-specific distance metrics for the comparison of plots. This paper describes a technique called LearnMet proposed to learn such metrics. The input to LearnMet is a training set with actual clusters of plots. These are iteratively compared with clusters over the same plots predicted using an arbitrary but fixed clustering algorithm. Using a guessed initial metric for clustering, adjustments are made to the metric in each epoch based on the error between the predicted and actual clusters until the error is minimal or below a given threshold. The metric giving the lowest error is output as the learned metric. The proposed LearnMet technique and its enhancements are discussed in detail in this paper. The primary application of LearnMet is clustering plots in the Heat Treating domain. Hence it is rigorously evaluated using Heat Treating data. Given distinct test sets for evaluation, clusters of plots predicted using the learned metrics are compared with given actual clusters over the same plots. The extent to which the predicted and actual clusters match each other denotes the accuracy of the learned metrics.

Original languageEnglish
Pages (from-to)29-53
Number of pages25
JournalMultimedia Tools and Applications
Volume35
Issue number1
DOIs
StatePublished - 1 Oct 2007

Fingerprint

Clustering algorithms
Semantics
Hot Temperature

Keywords

  • Clustering
  • Curve comparison
  • Distance metrics
  • Domain semantics
  • Parameter learning

Cite this

Varde, Aparna ; Rundensteiner, Elke ; Ruiz, Carolina ; Maniruzzaman, Mohammed ; Sisson, Richard. / LearnMet : Learning domain-specific distance metrics for plots of scientific functions. In: Multimedia Tools and Applications. 2007 ; Vol. 35, No. 1. pp. 29-53.
@article{2c31959642df40ab830215892c8d6a1c,
title = "LearnMet: Learning domain-specific distance metrics for plots of scientific functions",
abstract = "Scientific experimental results are often depicted as plots of functions to aid their visual analysis and comparison. In computationally comparing these plots using techniques such as similarity search and clustering, the notion of similarity is typically distance. However, it is seldom known which distance metric(s) best preserve(s) semantics in the respective domain. It is thus desirable to learn such domain-specific distance metrics for the comparison of plots. This paper describes a technique called LearnMet proposed to learn such metrics. The input to LearnMet is a training set with actual clusters of plots. These are iteratively compared with clusters over the same plots predicted using an arbitrary but fixed clustering algorithm. Using a guessed initial metric for clustering, adjustments are made to the metric in each epoch based on the error between the predicted and actual clusters until the error is minimal or below a given threshold. The metric giving the lowest error is output as the learned metric. The proposed LearnMet technique and its enhancements are discussed in detail in this paper. The primary application of LearnMet is clustering plots in the Heat Treating domain. Hence it is rigorously evaluated using Heat Treating data. Given distinct test sets for evaluation, clusters of plots predicted using the learned metrics are compared with given actual clusters over the same plots. The extent to which the predicted and actual clusters match each other denotes the accuracy of the learned metrics.",
keywords = "Clustering, Curve comparison, Distance metrics, Domain semantics, Parameter learning",
author = "Aparna Varde and Elke Rundensteiner and Carolina Ruiz and Mohammed Maniruzzaman and Richard Sisson",
year = "2007",
month = "10",
day = "1",
doi = "10.1007/s11042-007-0120-0",
language = "English",
volume = "35",
pages = "29--53",
journal = "Multimedia Tools and Applications",
issn = "1380-7501",
publisher = "Springer Netherlands",
number = "1",

}

LearnMet : Learning domain-specific distance metrics for plots of scientific functions. / Varde, Aparna; Rundensteiner, Elke; Ruiz, Carolina; Maniruzzaman, Mohammed; Sisson, Richard.

In: Multimedia Tools and Applications, Vol. 35, No. 1, 01.10.2007, p. 29-53.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - LearnMet

T2 - Learning domain-specific distance metrics for plots of scientific functions

AU - Varde, Aparna

AU - Rundensteiner, Elke

AU - Ruiz, Carolina

AU - Maniruzzaman, Mohammed

AU - Sisson, Richard

PY - 2007/10/1

Y1 - 2007/10/1

N2 - Scientific experimental results are often depicted as plots of functions to aid their visual analysis and comparison. In computationally comparing these plots using techniques such as similarity search and clustering, the notion of similarity is typically distance. However, it is seldom known which distance metric(s) best preserve(s) semantics in the respective domain. It is thus desirable to learn such domain-specific distance metrics for the comparison of plots. This paper describes a technique called LearnMet proposed to learn such metrics. The input to LearnMet is a training set with actual clusters of plots. These are iteratively compared with clusters over the same plots predicted using an arbitrary but fixed clustering algorithm. Using a guessed initial metric for clustering, adjustments are made to the metric in each epoch based on the error between the predicted and actual clusters until the error is minimal or below a given threshold. The metric giving the lowest error is output as the learned metric. The proposed LearnMet technique and its enhancements are discussed in detail in this paper. The primary application of LearnMet is clustering plots in the Heat Treating domain. Hence it is rigorously evaluated using Heat Treating data. Given distinct test sets for evaluation, clusters of plots predicted using the learned metrics are compared with given actual clusters over the same plots. The extent to which the predicted and actual clusters match each other denotes the accuracy of the learned metrics.

AB - Scientific experimental results are often depicted as plots of functions to aid their visual analysis and comparison. In computationally comparing these plots using techniques such as similarity search and clustering, the notion of similarity is typically distance. However, it is seldom known which distance metric(s) best preserve(s) semantics in the respective domain. It is thus desirable to learn such domain-specific distance metrics for the comparison of plots. This paper describes a technique called LearnMet proposed to learn such metrics. The input to LearnMet is a training set with actual clusters of plots. These are iteratively compared with clusters over the same plots predicted using an arbitrary but fixed clustering algorithm. Using a guessed initial metric for clustering, adjustments are made to the metric in each epoch based on the error between the predicted and actual clusters until the error is minimal or below a given threshold. The metric giving the lowest error is output as the learned metric. The proposed LearnMet technique and its enhancements are discussed in detail in this paper. The primary application of LearnMet is clustering plots in the Heat Treating domain. Hence it is rigorously evaluated using Heat Treating data. Given distinct test sets for evaluation, clusters of plots predicted using the learned metrics are compared with given actual clusters over the same plots. The extent to which the predicted and actual clusters match each other denotes the accuracy of the learned metrics.

KW - Clustering

KW - Curve comparison

KW - Distance metrics

KW - Domain semantics

KW - Parameter learning

UR - http://www.scopus.com/inward/record.url?scp=34548593322&partnerID=8YFLogxK

U2 - 10.1007/s11042-007-0120-0

DO - 10.1007/s11042-007-0120-0

M3 - Article

VL - 35

SP - 29

EP - 53

JO - Multimedia Tools and Applications

JF - Multimedia Tools and Applications

SN - 1380-7501

IS - 1

ER -