Multi–Criteria Decision Analysis: Methods and Soft ware

Alessio Ishizaka Reader in Decision Analysis, Portsmouth Business School, University of Portsmouth, UK Philippe Nemery Senior Research Scientist, SAP Labs - China, Shanghai, PRC

Foreword xi Acknowledgements xiii 1 General introduction 1 1.1 Introduction 1 1.2 Decision problems 3 1.3 MCDA methods 4 1.4 MCDA software 5 1.5 Selection of MCDA methods 5 1.6 Outline of the book 8 References 9 Part I FULL AGGREGATION APPROACH 11 2 Analytic hierarchy process 13 2.1 Introduction 13 2.2 Essential concepts of AHP 13 2.2.1 Problem structuring 14 2.2.2 Priority calculation 16 2.2.3 Consistency check 18 2.2.4 Sensitivity analysis 19 2.3 AHP software: MakeItRational 20 2.3.1 Problem structuring 20 2.3.2 Preferences and priority calculation 21 2.3.3 Consistency check 22 2.3.4 Results 24 2.3.5 Sensitivity analysis 25 2.4 In the black box of AHP 27 2.4.1 Problem structuring 27 2.4.2 Judgement scales 28 2.4.3 Consistency 31 2.4.4 Priorities derivation 33 2.4.5 Aggregation 39 2.5 Extensions of AHP 40 2.5.1 Analytic hierarchy process ordering 41 2.5.2 Group analytic hierarchy process 44 2.5.3 Clusters and pivots for a large number of alternatives 48 2.5.4 AHPSort 50 References 54 3 Analytic network process 59 3.1 Introduction 59 3.2 Essential concepts of ANP 59 3.2.1 Inner dependency in the criteria cluster 60 3.2.2 Inner dependency in the alternative cluster 63 3.2.3 Outer dependency 64 3.2.4 Influence matrix 67 3.3 ANP software: Super Decisions 68 3.3.1 Problem structuring 69 3.3.2 Assessment of pairwise comparison 70 3.3.3 Results 73 3.3.4 Sensitivity analysis 74 3.4 In the black box of ANP 76 3.4.1 Markov chain 76 3.4.2 Supermatrix 78 References 80 4 Multi-attribute utility theory 81 4.1 Introduction 81 4.2 Essential concepts of MAUT 81 4.2.1 The additive model 83 4.3 RightChoice 89 4.3.1 Data input and utility functions 89 4.3.2 Results 93 4.3.3 Sensitivity analysis 94 4.3.4 Group decision and multi-scenario analysis 95 4.4 In the black box of MAUT 97 4.5 Extensions of the MAUT method 98 4.5.1 The UTA method 98 4.5.2 UTAGMS 105 4.5.3 GRIP 111 References 112 5 MACBETH 114 5.1 Introduction 114 5.2 Essential concepts of MACBETH 114 5.2.1 Problem structuring: Value tree 115 5.2.2 Score calculation 117 5.2.3 Incompatibility check 118 5.3 Software description: M-MACBETH 122 5.3.1 Problem structuring: Value tree 122 5.3.2 Evaluations and scores 122 5.3.3 Incompatibility check 125 5.3.4 Results 127 5.3.5 Sensitivity analysis 127 5.3.6 Robustness analysis 127 5.4 In the black box of MACBETH 131 5.4.1 LP-MACBETH 131 5.4.2 Discussion 133 References 133 Part II OUTRANKING APPROACH 135 6 PROMETHEE 137 6.1 Introduction 137 6.2 Essential concepts of the PROMETHEE method 137 6.2.1 Unicriterion preference degrees 138 6.2.2 Unicriterion positive, negative and net flows 142 6.2.3 Global flows 143 6.2.4 The Gaia plane 146 6.2.5 Sensitivity analysis 148 6.3 The Smart Picker Pro software 149 6.3.1 Data entry 149 6.3.2 Entering preference parameters 151 6.3.3 Weights 153 6.3.4 PROMETHEE II ranking 155 6.3.5 Gaia plane 157 6.3.6 Sensitivity analysis 158 6.4 In the black box of PROMETHEE 160 6.4.1 Unicriterion preference degrees 162 6.4.2 Global preference degree 163 6.4.3 Global flows 164 6.4.4 PROMETHEE I and PROMETHEE II ranking 166 6.4.5 The Gaia plane 167 6.4.6 Influence of pairwise comparisons 168 6.5 Extensions of PROMETHEE 170 6.5.1 PROMETHEE GDSS 170 6.5.2 FlowSort: A sorting or supervised classification method 172 References 177 7 ELECTRE 180 7.1 Introduction 180 7.2 Essentials of the ELECTRE methods 180 7.2.1 ELECTRE III 183 7.3 The Electre III-IV software 189 7.3.1 Data entry 190 7.3.2 Entering preference parameters 191 7.3.3 Results 193 7.4 In the black box of ELECTRE III 194 7.4.1 Outranking relations 194 7.4.2 Partial concordance degree 195 7.4.3 Global concordance degree 196 7.4.4 Partial discordance degree 196 7.4.5 Outranking degree 197 7.4.6 Partial ranking: Exploitation of the outranking relations 199 7.4.7 Some properties 203 7.5 ELECTRE-Tri 204 7.5.1 Introduction 204 7.5.2 Preference relations 205 7.5.3 Assignment rules 207 7.5.4 Properties 207 References 210 Part III GOAL, ASPIRATION OR REFERENCE-LEVEL APPROACH 213 8 TOPSIS 215 8.1 Introduction 215 8.2 Essentials of TOPSIS 215 References 221 9 Goal programming 222 9.1 Introduction 222 9.2 Essential concepts of goal programming 222 9.3 Software description 227 9.3.1 Microsoft Excel Solver 227 9.4 Extensions of the goal programming 228 9.4.1 Weighted goal programming 228 9.4.2 Lexicographic goal programming 230 9.4.3 Chebyshev goal programming 232 References 234 10 Data Envelopment Analysis 235 Jean-Marc Huguenin 10.1 Introduction 235 10.2 Essential concepts of DEA 236 10.2.1 An efficiency measurement method 236 10.2.2 A DEA case study 237 10.2.3 Multiple outputs and inputs 247 10.2.4 Types of efficiency 248 10.2.5 Managerial implications 249 10.3 The DEA software 252 10.3.1 Building a spreadsheet in Win4DEAP 254 10.3.2 Running a DEA model 255 10.3.3 Interpreting results 257 10.4 In the black box of DEA 262 10.4.1 Constant returns to scale 263 10.4.2 Variable returns to scale 266 10.5 Extensions of DEA 268 10.5.1 Adjusting for the environment 268 10.5.2 Preferences 268 10.5.3 Sensitivity analysis 269 10.5.4 Time series data 270 References 270 Part IV INTEGRATED SYSTEMS 275 11 Multi-method platforms 277 11.1 Introduction 277 11.2 Decision Deck 278 11.3 DECERNS 278 11.3.1 The GIS module 279 11.3.2 The MCDA module 281 11.3.3 The GDSS module 284 11.3.4 Integration 286 References 287 Appendix: Linear optimization 288 A.1 Problem modelling 288 A.2 Graphical solution 289 A.3 Solution with Microsoft Excel 289 Index 293