Comparing mathematical and heuristic approaches for scientific data analysis

Aparna Varde, Shuhui Ma, Mohammed Maniruzzaman, David C. Brown, Elke A. Rundensteiner, Richard D. Sisson

Research output: Contribution to journalArticleResearchpeer-review

1 Citation (Scopus)

Abstract

Scientific data is often analyzed in the context of domain-specific problems, for example, failure diagnostics, predictive analysis, and computational estimation. These problems can be solved using approaches such as mathematical models or heuristic methods. In this paper we compare a heuristic approach based on mining stored data with a mathematical approach based on applying state-of-the-art formulae to solve an estimation problem. The goal is to estimate results of scientific experiments given their input conditions. We present a comparative study based on sample space, time complexity, and data storage with respect to a real application in materials science. Performance evaluation with real materials science data is also presented, taking into account accuracy and efficiency. We find that both approaches have their pros and cons in computational estimation. Similar arguments can be applied to other scientific problems such as failure diagnostics and predictive analysis. In the estimation problem in this paper, heuristic methods outperform mathematical models.

Original languageEnglish
Pages (from-to)53-69
Number of pages17
JournalArtificial Intelligence for Engineering Design, Analysis and Manufacturing: AIEDAM
Volume22
Issue number1
DOIs
StatePublished - 1 May 2008

Fingerprint

Heuristic methods
Materials science
Mathematical models
Data mining
Data storage equipment
Experiments
Predictive analytics

Keywords

  • Comparative study
  • Computational estimation
  • Heat treating of materials
  • Heuristic methods
  • Mathematical modeling

Cite this

Varde, Aparna ; Ma, Shuhui ; Maniruzzaman, Mohammed ; Brown, David C. ; Rundensteiner, Elke A. ; Sisson, Richard D. / Comparing mathematical and heuristic approaches for scientific data analysis. In: Artificial Intelligence for Engineering Design, Analysis and Manufacturing: AIEDAM. 2008 ; Vol. 22, No. 1. pp. 53-69.
@article{7b0e141af5684cb7a6a599c4292784ca,
title = "Comparing mathematical and heuristic approaches for scientific data analysis",
abstract = "Scientific data is often analyzed in the context of domain-specific problems, for example, failure diagnostics, predictive analysis, and computational estimation. These problems can be solved using approaches such as mathematical models or heuristic methods. In this paper we compare a heuristic approach based on mining stored data with a mathematical approach based on applying state-of-the-art formulae to solve an estimation problem. The goal is to estimate results of scientific experiments given their input conditions. We present a comparative study based on sample space, time complexity, and data storage with respect to a real application in materials science. Performance evaluation with real materials science data is also presented, taking into account accuracy and efficiency. We find that both approaches have their pros and cons in computational estimation. Similar arguments can be applied to other scientific problems such as failure diagnostics and predictive analysis. In the estimation problem in this paper, heuristic methods outperform mathematical models.",
keywords = "Comparative study, Computational estimation, Heat treating of materials, Heuristic methods, Mathematical modeling",
author = "Aparna Varde and Shuhui Ma and Mohammed Maniruzzaman and Brown, {David C.} and Rundensteiner, {Elke A.} and Sisson, {Richard D.}",
year = "2008",
month = "5",
day = "1",
doi = "10.1017/S0890060408000048",
language = "English",
volume = "22",
pages = "53--69",
journal = "Artificial Intelligence for Engineering Design, Analysis and Manufacturing: AIEDAM",
issn = "0890-0604",
publisher = "Cambridge University Press",
number = "1",

}

Comparing mathematical and heuristic approaches for scientific data analysis. / Varde, Aparna; Ma, Shuhui; Maniruzzaman, Mohammed; Brown, David C.; Rundensteiner, Elke A.; Sisson, Richard D.

In: Artificial Intelligence for Engineering Design, Analysis and Manufacturing: AIEDAM, Vol. 22, No. 1, 01.05.2008, p. 53-69.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Comparing mathematical and heuristic approaches for scientific data analysis

AU - Varde, Aparna

AU - Ma, Shuhui

AU - Maniruzzaman, Mohammed

AU - Brown, David C.

AU - Rundensteiner, Elke A.

AU - Sisson, Richard D.

PY - 2008/5/1

Y1 - 2008/5/1

N2 - Scientific data is often analyzed in the context of domain-specific problems, for example, failure diagnostics, predictive analysis, and computational estimation. These problems can be solved using approaches such as mathematical models or heuristic methods. In this paper we compare a heuristic approach based on mining stored data with a mathematical approach based on applying state-of-the-art formulae to solve an estimation problem. The goal is to estimate results of scientific experiments given their input conditions. We present a comparative study based on sample space, time complexity, and data storage with respect to a real application in materials science. Performance evaluation with real materials science data is also presented, taking into account accuracy and efficiency. We find that both approaches have their pros and cons in computational estimation. Similar arguments can be applied to other scientific problems such as failure diagnostics and predictive analysis. In the estimation problem in this paper, heuristic methods outperform mathematical models.

AB - Scientific data is often analyzed in the context of domain-specific problems, for example, failure diagnostics, predictive analysis, and computational estimation. These problems can be solved using approaches such as mathematical models or heuristic methods. In this paper we compare a heuristic approach based on mining stored data with a mathematical approach based on applying state-of-the-art formulae to solve an estimation problem. The goal is to estimate results of scientific experiments given their input conditions. We present a comparative study based on sample space, time complexity, and data storage with respect to a real application in materials science. Performance evaluation with real materials science data is also presented, taking into account accuracy and efficiency. We find that both approaches have their pros and cons in computational estimation. Similar arguments can be applied to other scientific problems such as failure diagnostics and predictive analysis. In the estimation problem in this paper, heuristic methods outperform mathematical models.

KW - Comparative study

KW - Computational estimation

KW - Heat treating of materials

KW - Heuristic methods

KW - Mathematical modeling

UR - http://www.scopus.com/inward/record.url?scp=37249035601&partnerID=8YFLogxK

U2 - 10.1017/S0890060408000048

DO - 10.1017/S0890060408000048

M3 - Article

VL - 22

SP - 53

EP - 69

JO - Artificial Intelligence for Engineering Design, Analysis and Manufacturing: AIEDAM

JF - Artificial Intelligence for Engineering Design, Analysis and Manufacturing: AIEDAM

SN - 0890-0604

IS - 1

ER -