A comparative study of multiresponse optimization criteria working ability

Nuno Costa; João Lourenço

doi:10.1016/j.chemolab.2014.08.004

A comparative study of multiresponse optimization criteria working ability

Nuno Costa, João Lourenço

Research output: Contribution to journal › Article › peer-review

13 Citations (Scopus)

Abstract

There is not a unique optimal solution for problems with multiple responses. Some solutions lead to operation conditions that are more hazardous, more costly or more difficult to implement and control. Therefore, it is useful for the analyst or decision-maker to use a criterion that can capture solutions evenly distributed along the so-called Pareto frontier. To provide information about criteria's working ability to depict Pareto frontiers, four optimization criteria built on different approaches were evaluated. Results show differences in criteria's performance. In particular, a consistent performance of a global criterion and limitations of the widely used desirability-based criterion is stated. (C) 2014 Elsevier B.V. All rights reserved.

Original language	English
Pages (from-to)	171-177
Number of pages	7
Journal	Chemometrics And Intelligent Laboratory Systems
Volume	138
DOIs	https://doi.org/10.1016/j.chemolab.2014.08.004
Publication status	Published - 15 Nov 2014

Keywords

Compromise
Desirability
Multiobjective
Optimal
Pareto
RSM
RESPONSE-SURFACE METHODOLOGY
PREFERENCE ARTICULATION APPROACH
PRINCIPAL COMPONENT ANALYSIS
ROBUST DESIGN OPTIMIZATION
DATA ENVELOPMENT ANALYSIS
MULTIOBJECTIVE OPTIMIZATION
DESIRABILITY FUNCTION
MULTIPLE RESPONSES
MULTICRITERIA OPTIMIZATION
MULTIVARIATE OPTIMIZATION

Access to Document

10.1016/j.chemolab.2014.08.004

Cite this

@article{12c8fa857d804ef38491de01c95b9335,

title = "A comparative study of multiresponse optimization criteria working ability",

abstract = "There is not a unique optimal solution for problems with multiple responses. Some solutions lead to operation conditions that are more hazardous, more costly or more difficult to implement and control. Therefore, it is useful for the analyst or decision-maker to use a criterion that can capture solutions evenly distributed along the so-called Pareto frontier. To provide information about criteria's working ability to depict Pareto frontiers, four optimization criteria built on different approaches were evaluated. Results show differences in criteria's performance. In particular, a consistent performance of a global criterion and limitations of the widely used desirability-based criterion is stated. (C) 2014 Elsevier B.V. All rights reserved.",

keywords = "Compromise, Desirability, Multiobjective, Optimal, Pareto, RSM, RESPONSE-SURFACE METHODOLOGY, PREFERENCE ARTICULATION APPROACH, PRINCIPAL COMPONENT ANALYSIS, ROBUST DESIGN OPTIMIZATION, DATA ENVELOPMENT ANALYSIS, MULTIOBJECTIVE OPTIMIZATION, DESIRABILITY FUNCTION, MULTIPLE RESPONSES, MULTICRITERIA OPTIMIZATION, MULTIVARIATE OPTIMIZATION",

author = "Nuno Costa and Jo{\~a}o Louren{\c c}o",

year = "2014",

month = nov,

day = "15",

doi = "10.1016/j.chemolab.2014.08.004",

language = "English",

volume = "138",

pages = "171--177",

journal = "Chemometrics And Intelligent Laboratory Systems",

issn = "0169-7439",

publisher = "Elsevier Science B.V., Inc",

}

TY - JOUR

T1 - A comparative study of multiresponse optimization criteria working ability

AU - Costa, Nuno

AU - Lourenço, João

PY - 2014/11/15

Y1 - 2014/11/15

N2 - There is not a unique optimal solution for problems with multiple responses. Some solutions lead to operation conditions that are more hazardous, more costly or more difficult to implement and control. Therefore, it is useful for the analyst or decision-maker to use a criterion that can capture solutions evenly distributed along the so-called Pareto frontier. To provide information about criteria's working ability to depict Pareto frontiers, four optimization criteria built on different approaches were evaluated. Results show differences in criteria's performance. In particular, a consistent performance of a global criterion and limitations of the widely used desirability-based criterion is stated. (C) 2014 Elsevier B.V. All rights reserved.

AB - There is not a unique optimal solution for problems with multiple responses. Some solutions lead to operation conditions that are more hazardous, more costly or more difficult to implement and control. Therefore, it is useful for the analyst or decision-maker to use a criterion that can capture solutions evenly distributed along the so-called Pareto frontier. To provide information about criteria's working ability to depict Pareto frontiers, four optimization criteria built on different approaches were evaluated. Results show differences in criteria's performance. In particular, a consistent performance of a global criterion and limitations of the widely used desirability-based criterion is stated. (C) 2014 Elsevier B.V. All rights reserved.

KW - Compromise

KW - Desirability

KW - Multiobjective

KW - Optimal

KW - Pareto

KW - RSM

KW - RESPONSE-SURFACE METHODOLOGY

KW - PREFERENCE ARTICULATION APPROACH

KW - PRINCIPAL COMPONENT ANALYSIS

KW - ROBUST DESIGN OPTIMIZATION

KW - DATA ENVELOPMENT ANALYSIS

KW - MULTIOBJECTIVE OPTIMIZATION

KW - DESIRABILITY FUNCTION

KW - MULTIPLE RESPONSES

KW - MULTICRITERIA OPTIMIZATION

KW - MULTIVARIATE OPTIMIZATION

U2 - 10.1016/j.chemolab.2014.08.004

DO - 10.1016/j.chemolab.2014.08.004

M3 - Article

SN - 0169-7439

VL - 138

SP - 171

EP - 177

JO - Chemometrics And Intelligent Laboratory Systems

JF - Chemometrics And Intelligent Laboratory Systems

ER -