Multi-Criteria Analysis - preference weighting. Defining weights for criteria Purpose: to express the importance of each criterion relative to other criteria.

Slides:



Advertisements
Similar presentations
Modellistica e Gestione dei Sistemi Ambientali A tool for multicriteria analysis: The Analytic Hierarchy Process Chiara Mocenni University of.
Advertisements

DECISION MODELING WITH Multi-Objective Decision Making
Multi‑Criteria Decision Making
Analytical Hierarchy Process (AHP) - by Saaty
1 1 Slide Chapter 10 Multicriteria Decision Making n A Scoring Model for Job Selection n Spreadsheet Solution of the Job Selection Scoring Model n The.
By Joseph K. Berry W. M. Keck Scholar, University of Denver – August, An Overview of GIS-based Corridor.
MIS 463 Analytic Hierarchy Process. 2 The Analytic Hierarchy Process (AHP) It is popular and widely used method for multi-criteria decision making. Allows.
Lecture 08 Analytic Hierarchy Process (Module 1)
1 Critical Success Factors and Organizational Performance Prepared by: Niemann, Lahlou, Zertani & Pflug Lecturer: Ihsan Yüksel.
Introduction to Management Science
Raster Based GIS Analysis
Copyright © 2006 Pearson Education Canada Inc Course Arrangement !!! Nov. 22,Tuesday Last Class Nov. 23,WednesdayQuiz 5 Nov. 25, FridayTutorial 5.
Grid-based GIS Modeling Nigel Trodd Modified from Berry JK, GIS Modeling, presented at Grid-based Map Analysis Techniques and Modeling Workshop,
1 1 Slide © 2005 Thomson/South-Western EMGT 501 HW Solutions Problem Problem
Multi Criteria Decision Modeling Preference Ranking The Analytical Hierarchy Process.
1 The Analytic Hierarchy Process. 2 Overview of the AHP 1.Set up decision hierarchy 2.Make pairwise comparisons of attributes and alternatives 3.Transform.
Assignment of Weights Other methods, besides arbitrary, for weight assignment exist There are both direct and indirect weight elicitation techniques Source:
Executive Manager Decision Making and Policy Planning, typically with many goals Sometimes even > 1 decision maker (Game Theory, Group Decisions) Linear.
MENENTUKAN LOKASI PABRIK YANG IDEAL MENGGUNAKAN AHP PERTEMUAN 12.
Introduction to Management Science
1 1 Slide © 2005 Thomson/South-Western EMGT 501 HW Solutions Chapter 14 - SELF TEST 20.
Strategic Project Alignment With Team Expert Choice
9-1 Copyright © 2010 Pearson Education, Inc. Publishing as Prentice Hall Multicriteria Decision Making Chapter 9.
Multicriteria Decision Making
9-1 Copyright © 2013 Pearson Education, Inc. Publishing as Prentice Hall Multicriteria Decision Making Chapter 9.
«Enhance of ship safety based on maintenance strategies by applying of Analytic Hierarchy Process» DAGKINIS IOANNIS, Dr. NIKITAKOS NIKITAS University of.
Presented by Johanna Lind and Anna Schurba Facility Location Planning using the Analytic Hierarchy Process Specialisation Seminar „Facility Location Planning“
Analytical Hierarchy Process ( AHP )
Quantitative Analysis for Management Multifactor Evaluation Process and Analytic Hierarchy Process Dr. Mohammad T. Isaai Graduate School of Management.
1 1 Slide © 2004 Thomson/South-Western Chapter 17 Multicriteria Decisions n Goal Programming n Goal Programming: Formulation and Graphical Solution and.
The Decision Making Process with EC2000-Keypad and Internet Versions.
Spreadsheet Modeling and Decision Analysis, 3e, by Cliff Ragsdale. © 2001 South-Western/Thomson Learning Multicriteria Decision Making u Decision.
ArcGIS Spatial Analyst – Suitability Modeling
Decision Technology Modeling, Software and Applications Matthew J. Liberatore Robert L. Nydick John Wiley & Sons, Inc.
Multi-Criteria Decision Making by: Mehrdad ghafoori Saber seyyed ali
1 Chapter 16 The Analytic Hierarchy Process. 2 The analytic hierarchy process (AHP), which was developed by Thomas Saaty when he was acting as an adviser.
Recap: How the Process Works (1) Determine the weights. The weights can be absolute or relative. Weights encompass two parts -- the quantitative weight.
Chapter 9 - Multicriteria Decision Making 1 Chapter 9 Multicriteria Decision Making Introduction to Management Science 8th Edition by Bernard W. Taylor.
MAINTENANCE STRATEGY SELECTION BASED ON HYBRID AHP-GP MODEL SUZANA SAVIĆ GORAN JANAĆKOVIĆ MIOMIR STANKOVIĆ University of Niš, Faculty of Occupational Safety.
Agenda for This Week Wednesday, April 27 AHP Friday, April 29 AHP Monday, May 2 Exam 2.
THE ANALYTIC HIERARCHY PROCESS CAR PURCHASE EXAMPLE.
Analytic Hierarchy Process. 2 The Analytic Hierarchy Process (AHP) Founded by Saaty in It is a popular and widely used method for multi-criteria.
To accompany Quantitative Analysis for Management, 9e \by Render/Stair/Hanna M1-1 © 2006 by Prentice Hall, Inc. Upper Saddle River, NJ Analytic Hierarchy.
BUSINESS PERFORMANCE MANAGEMENT
Esri UC 2014 | Technical Workshop | ArcGIS Spatial Analyst – Suitability Modeling Kevin M. Johnston Elizabeth Graham.
Analytic Hierarchy Process (AHP)
Pairwise-Comparison was used as the ranking method which provided accurate weights in determining the ideal site for Niagara Region. Site Selection for.
« Investment scenarios and regional factors in the solar energy sector » Dr. Nikolaos Apostolopoulos The Macrojournals /MacroTrends Conference: New York.
Model Calibration and Weighting Avoid areas of… High Housing Density Far from Roads In or Near Sensitive Areas High Visual Exposure …what is “high” housing.
Applied Mathematics 1 Applications of the Multi-Weighted Scoring Model and the Analytical Hierarchy Process for the Appraisal and Evaluation of Suppliers.
Constructing the PAHP-based Decision Support System by Considering the Ambiguity in Decision Making Norihiro Saikawa Department of Computer and Information.
ON ELICITATION TECHNIQUES OF NEAR-CONSISTENT PAIRWISE COMPARISON MATRICES József Temesi Department of Operations Research Corvinus University of Budapest,
ESTIMATING WEIGHT Course: Special Topics in Remote Sensing & GIS Mirza Muhammad Waqar Contact: EXT:2257 RG712.
MCE: Eigen Values Calculations from Pair Wise Comparisons. Addition to Exercise 2-8.
This Briefing is: UNCLASSIFIED Aha! Analytics 2278 Baldwin Drive Phone: (937) , FAX: (866) An Overview of the Analytic Hierarchy Process.
Analysis Manager Training Module
Reality of Highway Construction Equipment in Palestine
Supplement S7 Supplier Selection.
MULTI CRITERIA DECISION MAKING - APPLICATIONS IN PROJECT MANAGEMENT
Analytic Hierarchy Process (AHP)
A Scoring Model for Job Selection
ANALYTIC HIERARCHY PROCESS (AHP)
The Decision Making Process with EC2000-Keypad and Internet Versions
Analytic Hierarchy Process Prepared by Lee Revere and John Large
Analytic Hierarchy Process (AHP)
Slides by John Loucks St. Edward’s University.
Agenda for This Week Monday, April 25 AHP Wednesday, April 27
Conjoint analysis.
Multicriteria Decision Making
IME634: Management Decision Analysis
Presentation transcript:

Multi-Criteria Analysis - preference weighting

Defining weights for criteria Purpose: to express the importance of each criterion relative to other criteria

Overview Preference weighting techniques –Point allocation –Ranking methods –Pairwise comparison Multi-criteria analysis Compromise programming ArcGIS ranking model extension

Ranking or prioritization studies 1.Define the problem or goal 2.Decide on evaluation criteria 3.Normalize and create utility scales 4.Define weights for criteria 5.Calculate a ranking model result 6.Evaluate result and perform sensitivity analysis

Weighting methods Differ in: –Accuracy –Ease of use –Understanding of decision makers –Theoretical foundation

Weighting A value assigned to an evaluation criterion that indicates its importance relative (not absolute) to other criteria under consideration (Malczewski, 1999) Weights are normalized to be within a range of 0 to 1

Methods to create weights for criteria A. Rating methods –Point allocation B. Ranking methods –Rank sum –Rank reciprocal –Rank exponent C. Pairwise comparison –AHP

A. Rating methods Require the decision maker to estimate weights on the basis of a predetermined scale ie. 0 to = the criterion should not be considered 100 = the criterion is the only one that should be considered

Point allocation Weights are estimated by the decision maker on a predetermined scale, for example 0 to 100. In this approach, the more points a criterion receives, the greater its relative importance. The total of all criterion weights must sum to 100. This method is easy to normalize. Weight Elev change10 Wetlands25 Roads5 Row crops40 Mining Total100 Criteria that are important for ranking watersheds for runoff potential

Point allocation method + unlimited number of judgments + very easy to use + applicable to GIS applications -No underlying theoretical foundation -Not very precise -Difficult with more than 6 to 8 criteria

B. Ranking methods The easiest method for decision makers because they only need to come up with a straight rank Most important criterion= 1, next = 2, next = 3, and so on….. Once straight rank is done, the analyst has methods to derive weights

Rank sum, reciprocal, exponent 1 = most important criterion 5 = least important of the five criterion Straight rank is first step

Rank sum, reciprocal, exponent n = total number of criteria r j = the straight rank Straight rank is first step

Ranking methods + a good first step in working with decision makers toward consensus + spreadsheet calculation + applicable to GIS applications -Becomes difficult to straight rank many criteria -No underlying theoretical foundation - Not very precise, just approximations

C. Pairwise comparison Used by NRCS, USFS, Park Service in land management Developed by psychoanalysts

Background on pairwise comparison A well-developed method of ordering attributes or characteristics of a given set of items (Saaty, 1972) – AHP analytical hierarchy process Every response the individual makes is reduced to a simple choice between two items. The result of these paired choices is a set of items, rank-ordered in terms of which items are chosen more often.

Other notes Using the method of paired comparisons can be helpful for ranking items that are too numerous or too similar to mentally rank order. This technique is objective, simple, and also ensures that every participant in a workgroup or public meeting has a voice in how things get prioritized.

Pairwise comparison

Prefer Equal Somewhat prefer Strongly prefer

Example of pc sheet

Pairwise comparison limitation Number of full pairwise comparisons add up quickly! N = number of judgments Comparisons = (N(N – 1)) / 2 For 10 criteria or judgments….. comparisons = (10(10-1))/2 = 45

To limit full comparisons… Assume transitivity of preference (an abbreviated pairwise test) If A is preferred to B and B is preferred to C then we assume A is preferred to C

Ex of abbreviated pc sheet Abbreviated pairwise test does not allow for tests of consistency

Ex of full pc test

Step 1. create a comparison matrix For a trout suitability model, assume we use three criteria temperature (temp) water quality (wq) stream structure (Strc) CriterionTempWqStrc Temp Wq Strc

Step 2. Get comparisons using selected intensities 1 = equal preference 3 = prefer 6 = strongly prefer For finding suitable trout sites we decide that: Temp is preferred over Wq Temp is strongly preferred over Strc Wq is preferred over Strc

Step 3.a After making the diagonal values = 1, Fill in upper right corner of the matrix CriterionTempWqStrc Temp136 Wq13 Strc1 1 = equal preference 3 = prefer 6 = strongly prefer For finding suitable trout locations we decide that: Temp is preferred over Wq Tempo is strongly preferred over Strc Wq is preferred over Strc

Step 3.b Fill in matrix with reciprocal values CriterionTempWqStrc Temp136 Wq1/313 Strc1/61/31

Step 4.a Sum columns CriterionTempWqStrc Temp136 Wq1/313 Strc1/61/

Step 4.b Normalize the matrix CriterionTempWqStrc Temp136 Wq1/313 Strc1/61/ CriterionTempWqStrc Temp Wq Strc

Step 4.c Find the average of the rows, this will be the weights for the criteria CriterionTempWqStrc Temp Wq Strc ( ) / 3 = Temperature ( ) / 3 = Water quality ( ) / 3 = Stream structure Sum to 1

Step 5. Check for consistency in comparisons by first computing a consistency vector Criterion TempWqStrc Temp136 Wq Strc Temperature Water quality Stream struc Temp(0.653)(1) + (0.251)(3) + (0.096)(6) = Wq(0.653)(.333) + (0.251)(1) + (0.096)(3) = Strc(0.653)(.166) + (0.251)(.333) + (0.096)(1) = Original matrix from 3.b Weights from 4.c Temp1.982 / = Wq0.756 / = Strc0.287 / = New consistency vector

Step 6.a Now that we have the consistency vector, we need to compute two more terms, lambda (λ) and the consistency index (CI) λ = the average of the consistency vector from step 5. = ( ) / 3 = 3.101

Step 6.b Compute the consistency index which provides a measure from consistency –CI = (λ – N) / (N- 1) –N = number of criteria CI = – 3 / 3 -1 = 0.050

Step 7. With the consistency index (CI) we can now use the random index (RI) table to compute the consistency ratio (CR) CR = CI / RI Number of criteria RI Saaty (1999) p.186 = / 0.58 = 0.086

Step 8. Interpretation of the consistency ration (CR) If CR < then there is consistency in responses In our example, since the CR is which is < then we can use the weights from step 4c.! Temperature Water quality Stream structure

Step 9. Integrate into GIS the field calculator (vector table field) or the raster calculator (grid data)

Pairwise comparison + good theoretical foundation for further statistical test (Mann-Whitney) + easy for decision makers to understand + applicable to GIS applications - Becomes difficult to straight rank many criteria -Must check for consistency or assume transitivity of preferences -Time consuming computations -Path dependency issue

Summary Which method to use depends on the trade- offs one is willing to make between –Ease of use –Accuracy –Degree of understanding on the part of the decision maker –Theoretical foundation –Availability of computer software

FYI: Software for weighting Criterion Decision Plus Expert Choice Logical Decisions Pairwise comparison program by NRCS AHP Arcscripts toolbar for ArcGIS

Other options for weights Use modeled output from logistic regression or principal component analysis Past studies or published literature Interview experts Habitat suitabilty indices (HSIs) for wildlife studies

Questions / comments?