Group Aggregation Techniques for Analytic Hierarchy Process and Analytic Network Process: A Comparative Analysis

Ossadnik, Wolfgang; Schinke, Stefanie; Kaspar, Ralf

doi:10.1007/s10726-015-9448-4

Cited by 170 publications

(105 citation statements)

References 37 publications

Supporting

Mentioning

101

Contrasting

Unclassified

Order By: Relevance

“…Although the example we showed here only involved a single policy maker, our framework works for a group of policy makers as well. This is because the Analytic Hierarchy Process (AHP) has been extended to aggregate group preferences [15]. This makes our framework well suited for real world decision tasks where the stakeholders often have very different backgrounds and interests.…”

Section: Resultsmentioning

confidence: 99%

Joint Optimization of AI Fairness and Utility

Zhang

Bellamy

Varshney

2020

Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society

View full text Add to dashboard Cite

Today, AI is increasingly being used in many high-stakes decisionmaking applications in which fairness is an important concern. Already, there are many examples of AI being biased and making questionable and unfair decisions. The AI research community has proposed many methods to measure and mitigate unwanted biases, but few of them involve inputs from human policy makers. We argue that because different fairness criteria sometimes cannot be simultaneously satisfied, and because achieving fairness often requires sacrificing other objectives such as model accuracy, it is key to acquire and adhere to human policy makers' preferences on how to make the tradeoff among these objectives. In this paper, we propose a framework and some exemplar methods for eliciting such preferences and for optimizing an AI model according to these preferences.

show abstract

Section: Resultsmentioning

confidence: 99%

Joint Optimization of AI Fairness and Utility

Zhang

Bellamy

Varshney

2020

Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society

View full text Add to dashboard Cite

show abstract

“…Two scenarios were created for each expert: one with the AHP and the other with the ANP method. In addition, a group scenario was generated by aggregating individual priorities (AIP) using the geometric mean (Ossadnik et al, 2016). The resultant maps were classified into five categories of vulnerability to facilitate their interpretation and comparison: very low (0.00-0.20), low (0.20-0.40), medium (0.40-0.60), high (0.60-0.80), and very high (0.80-1.00).…”

Section: Aggregation Of Criteria To Create Flood Vulnerability Mapsmentioning

confidence: 99%

Participatory flood vulnerability assessment: a multi-criteria approach

Brito

Evers

Almoradie

2018

Hydrol. Earth Syst. Sci.

132

View full text Add to dashboard Cite

Abstract. This paper presents a participatory multi-criteria decision-making (MCDM) approach for flood vulnerability assessment while considering the relationships between vulnerability criteria. The applicability of the proposed framework is demonstrated in the municipalities of Lajeado and Estrela, Brazil. The model was co-constructed by 101 experts from governmental organizations, universities, research institutes, NGOs, and private companies. Participatory methods such as the Delphi survey, focus groups, and workshops were applied. A participatory problem structuration, in which the modellers work closely with end users, was used to establish the structure of the vulnerability index. The preferences of each participant regarding the criteria importance were spatially modelled through the analytical hierarchy process (AHP) and analytical network process (ANP) multi-criteria methods. Experts were also involved at the end of the modelling exercise for validation. The final product is a set of individual and group flood vulnerability maps. Both AHP and ANP proved to be effective for flood vulnerability assessment; however, ANP is preferred as it considers the dependences among criteria. The participatory approach enabled experts to learn from each other and acknowledge different perspectives towards social learning. The findings highlight that to enhance the credibility and deployment of model results, multiple viewpoints should be integrated without forcing consensus.

show abstract

“…(Butler et al, 1997;Chen & Kocaoglu, 2008;May, Shang, Tjader, & Vargas, 2013;Triantaphyllou & Sanchez, 1997) Group decision making Uncertainty associated with the combination procedure of several decision makers' judgments-How (e.g., geometric mean on pairwise comparisons, weighted arithmetic mean on derived priorities, or consensus models) to derive an appropriate group aggregation? (Altuzarra et al, 2007;Dong et al, 2010;Forman & Peniwati, 1998;Grošelj et al, 2015;Ishizaka & Labib, 2011a;Mikhailov, 2004b;Ossadnik et al, 2016;Saaty & Peniwati, 2013;Saaty & Vargas, 2012) Note. AHP: analytic hierarchy process; MCDM: multicriteria decision making.…”

Section: Embedded Uncertainty Issuesmentioning

confidence: 99%

A comprehensive uncertainty analysis of the analytic hierarchy process methodology applied in the context of environmental decision making

Toth

Vacik

2018

Multi Criteria Decision Anal

View full text Add to dashboard Cite

An overview of uncertainty issues associated with the analytic hierarchy process (AHP) is provided. Further, an explicit understanding of uncertainty (designation, categorization, and quantification) with respect to the methodological properties of the AHP is developed and used to analyse a hypothetical group decision problem located in the context of environmental decision making (EDM). To calculate the numerical impact of especially designed uncertainty scenarios (USs) on the final ranking given by the AHP, a simulation experiment is conducted using R. It evaluates the impact of uncertainty within three variants of a hypothetical decision-making case by calculating an "overall uncertainty" measure. The consideration of uncertainty may lead to a rank reversal in comparison with that analysis neglecting uncertainty (best alternative given by the AHP). The results show that the absolute maximal impact caused by a US is approximately 0.03. With respect to a single US and the specific case characteristics, in about 50% of the simulated runs a rank reversal occurs. Similar shares of rank reversal over different USs within a single variant of the case raise the question to which uncertainty should be given prior attention in decision-making practice. For decision analysts in EDM, this result implies that additional resources may be necessary to commonly negotiate with decision makers that uncertainties should be addressed. From a theoretical normative point of view, the effects of considering uncertainty issues in the AHP methodology cannot satisfy the ideal of a rational decision analysis. From a descriptive point of view, considering the practice of decision makers, the impacts of the considered uncertainties stay within reasonable limits, meaning that the maximal numerical impact stays on the hundredths decimal place. KEYWORDS designation, categorization, and quantification of uncertainty, overall uncertainty measure, scenarios, simulation experiment, uncertainty modellingThis is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

show abstract

Group Aggregation Techniques for Analytic Hierarchy Process and Analytic Network Process: A Comparative Analysis

Cited by 170 publications

References 37 publications

Joint Optimization of AI Fairness and Utility

Joint Optimization of AI Fairness and Utility

Participatory flood vulnerability assessment: a multi-criteria approach

A comprehensive uncertainty analysis of the analytic hierarchy process methodology applied in the context of environmental decision making

Contact Info

Product

Resources

About