ebook img

Multicriteria Decision Aid and Artificial Intelligence PDF

353 Pages·2013·2.926 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Multicriteria Decision Aid and Artificial Intelligence

Multicriteria Decision Aid and Artificial Intelligence Multicriteria Decision Aid and Artificial Intelligence Links, Theory and Applications Edited by Michael Doumpos and Evangelos Grigoroudis Technical University of Crete, Greece A John Wiley & Sons, Ltd., Publication Thiseditionfirstpublished2013 2013JohnWiley&Sons,Ltd Registeredoffice JohnWiley&SonsLtd,TheAtrium,SouthernGate,Chichester,WestSussex,PO198SQ,UnitedKingdom Fordetailsofourglobaleditorialoffices,forcustomerservicesandforinformationabouthowtoapplyforpermission toreusethecopyrightmaterialinthisbookpleaseseeourwebsiteatwww.wiley.com. TherightoftheauthortobeidentifiedastheauthorofthisworkhasbeenassertedinaccordancewiththeCopyright, DesignsandPatentsAct1988. Allrightsreserved.Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,ortransmitted,inany formorbyanymeans,electronic,mechanical,photocopying,recordingorotherwise,exceptaspermittedbytheUK Copyright,DesignsandPatentsAct1988,withoutthepriorpermissionofthepublisher. Wileyalsopublishesitsbooksinavarietyofelectronicformats.Somecontentthatappearsinprintmaynotbe availableinelectronicbooks. Designationsusedbycompaniestodistinguishtheirproductsareoftenclaimedastrademarks.Allbrandnamesand productnamesusedinthisbookaretradenames,servicemarks,trademarksorregisteredtrademarksoftheirrespective owners.Thepublisherisnotassociatedwithanyproductorvendormentionedinthisbook.Thispublicationisdesigned toprovideaccurateandauthoritativeinformationinregardtothesubjectmattercovered.Itissoldontheunderstanding thatthepublisherisnotengagedinrenderingprofessionalservices.Ifprofessionaladviceorotherexpertassistanceis required,theservicesofacompetentprofessionalshouldbesought. LibraryofCongressCataloging-in-PublicationData Doumpos,Michael. Multicriteriadecisionaidandartificialintelligence:links,theoryandapplications/MichaelDoumpos, EvangelosGrigoroudis. p.cm. Includesbibliographicalreferencesandindex. ISBN978-1-119-97639-4(hardback) 1. Multiplecriteriadecisionmaking.2. Artificialintelligence. I. Grigoroudis,Evangelos. II. Title. T57.95.D5782013 658.4’033–dc23 2012040171 AcataloguerecordforthisbookisavailablefromtheBritishLibrary. ISBN:978-1-119-97639-4 Setin10/12ptTimesbyLaserwordsPrivateLimited,Chennai,India Contents Preface xi Notes on Contributors xv Part I THECONTRIBUTIONSOFINTELLIGENTTECHNIQUESIN MULTICRITERIA DECISION AIDING 1 1 Computational intelligence techniques for multicriteria decision aiding: An overview 3 Michael Doumpos and Constantin Zopounidis 1.1 Introduction 3 1.2 The MCDA paradigm 4 1.2.1 Modeling process 4 1.2.2 Methodological approaches 6 1.3 Computational intelligence in MCDA 9 1.3.1 Statistical learning and data mining 9 1.3.2 Fuzzy modeling 12 1.3.3 Metaheuristics 15 1.4 Conclusions 17 References 18 2 Intelligent decision support systems 25 Gloria Phillips-Wren 2.1 Introduction 25 2.2 Fundamentals of human decision making 26 2.3 Decision support systems 29 2.4 Intelligent decision support systems 30 2.4.1 Artificial neural networks for intelligent decision support 31 2.4.2 Fuzzy logic for intelligent decision support 34 2.4.3 Expert systems for intelligent decision support 35 vi CONTENTS 2.4.4 Evolutionary computing for intelligent decision support 35 2.4.5 Intelligent agents for intelligent decision support 36 2.5 Evaluating intelligent decision support systems 38 2.5.1 Determining evaluation criteria 38 2.5.2 Multi-criteria model for IDSS assessment 39 2.6 Summary and future trends 40 Acknowledgment 41 References 41 Part II INTELLIGENTTECHNOLOGIESFORDECISIONSUPPORT AND PREFERENCE MODELING 45 3 Designing distributed multi-criteria decision support systems for complex and uncertain situations 47 Tina Comes, Niek Wijngaards and Frank Schultmann 3.1 Introduction 47 3.2 Example applications 49 3.3 Key challenges 51 3.4 Making trade-offs: Multi-criteria decision analysis 53 3.4.1 Multi-attribute decision support 53 3.4.2 Making trade-offs under uncertainty 55 3.5 Exploring the future: Scenario-based reasoning 56 3.6 Making robust decisions: Combining MCDA and SBR 57 3.6.1 Decisions under uncertainty: The concept of robustness 57 3.6.2 Combining scenarios and MCDA 58 3.6.3 Collecting, sharing and processing information: A distributed approach 59 3.6.4 Keeping trackof future developments: Constructing comparable scenarios 61 3.6.5 Respecting constraints and requirements: Scenario management 64 3.6.6 Assisting evaluation: Assessing large numbers of scenarios 66 3.7 Discussion 69 3.8 Conclusion 71 Acknowledgment 71 References 72 4 Preference representation with ontologies 77 Aida Valls, Antonio Moreno and Joan Borra`s 4.1 Introduction 77 4.2 Ontology-based preference models 80 4.3 Maintaining the user profile up to date 85 4.4 Decision making methods exploiting the preference information stored in ontologies 88 4.4.1 Recommendation based on aggregation 91 4.4.2 Recommendation based on similarities 92 4.4.3 Recommendation based on rules 93 CONTENTS vii 4.5 Discussion and open questions 94 Acknowledgments 95 References 96 Part III DECISION MODELS 101 5 Neural networks in multicriteria decision support 103 Thomas Hanne 5.1 Introduction 103 5.2 Basic concepts of neural networks 104 5.2.1 Neural networks for intelligent decision support 109 5.3 Basics in multicriteria decision aid 111 5.3.1 MCDM problems 111 5.3.2 Solutions of MCDM problems 112 5.4 Neural networks and multicriteria decision support 113 5.4.1 Review of neural network applications to MCDM problems 115 5.4.2 Discussion 121 5.5 Summary and conclusions 122 References 123 6 Rule-based approach to multicriteria ranking 127 Marcin Szela¸g, Salvatore Greco and Roman Słowin´ski 6.1 Introduction 127 6.2 Problem setting 130 6.3 Pairwise comparison table 132 6.4 Rough approximation of outranking and nonoutranking relations 133 6.5 Induction and application of decision rules 136 6.6 Exploitation of preference graphs 139 6.7 Illustrative example 149 6.8 Summary and conclusions 155 Acknowledgment 155 References 155 Appendix 159 7 About the application of evidence theory in multicriteria decision aid 161 Mohamed Ayman Boujelben and Yves De Smet 7.1 Introduction 161 7.2 Evidence theory: Some concepts 163 7.2.1 Knowledge model 163 7.2.2 Combination 164 7.2.3 Decision making 165 7.3 New concepts in evidence theory for MCDA 165 7.3.1 First belief dominance 165 7.3.2 RBBD concept 167 7.4 Multicriteria methods modeled by evidence theory 169 7.4.1 Evidential reasoning approach 169 viii CONTENTS 7.4.2 DS/AHP 172 7.4.3 DISSET 174 7.4.4 A choice model inspired by ELECTRE I 176 7.4.5 A ranking model inspired by Xu et al.’s method 179 7.5 Discussion 181 7.6 Conclusion 183 References 183 Part IV MULTIOBJECTIVE OPTIMIZATION 189 8 Interactive approaches applied to multiobjective evolutionary algorithms 191 Antonio Lo´pez Jaimes and Carlos A. Coello Coello 8.1 Introduction 191 8.1.1 Methods analyzed in this chapter 192 8.2 Basic concepts and notation 193 8.2.1 Multiobjective optimization problems 193 8.2.2 Classical interactive methods 195 8.3 MOEAs based on reference point methods 196 8.3.1 A weighted distance metric 196 8.3.2 Light beam search combined with NSGA-II 198 8.3.3 Controlling the accuracy of the Pareto front approximation 198 8.3.4 Light beam search combined with PSO 199 8.3.5 A preference relation based on a weighted distance metric 199 8.3.6 The Chebyshev preference relation 200 8.4 MOEAs based on value function methods 202 8.4.1 Progressive approximation of a value function 202 8.4.2 Value function by ordinal regression 202 8.5 Miscellaneous methods 203 8.5.1 Desirability functions 203 8.6 Conclusions and future work 204 Acknowledgment 205 References 205 9 Generalized data envelopment analysis and computational intelligence in multiple criteria decision making 209 Yeboon Yun and Hirotaka Nakayama 9.1 Introduction 209 9.2 Generalized data envelopment analysis 211 9.2.1 Basic DEA models: CCR, BCC and FDH models 212 9.2.2 GDEA model 214 9.3 Generation of Pareto optimal solutions using GDEA and computational intelligence 217 9.3.1 GDEA in fitness evaluation 217 9.3.2 GDEA in deciding the parameters of multi-objective PSO 222 9.3.3 Expected improvement for multi-objective optimization using GDEA 225 CONTENTS ix 9.4 Summary 229 References 231 10 Fuzzy multiobjective optimization 235 Masatoshi Sakawa 10.1 Introduction 235 10.2 Solution concepts for multiobjective programming 236 10.3 Interactive multiobjective linear programming 237 10.4 Fuzzy multiobjective linear programming 241 10.5 Interactive fuzzy multiobjective linear programming 242 10.6 Interactive fuzzy multiobjective linear programming with fuzzy parameters 248 10.7 Interactive fuzzy stochastic multiobjective linear programming 257 10.8 Related works and applications 266 References 267 Part V APPLICATIONS IN MANAGEMENT AND ENGINEERING 273 11 Multiple criteria decision aid and agents: Supporting effective resource federation in virtual organizations 275 Pavlos Delias and Nikolaos Matsatsinis 11.1 Introduction 275 11.2 The intuition of MCDA in multi-agent systems 276 11.3 Resource federation applied 277 11.3.1 Describing the problem in a cloud computing context 277 11.3.2 Problem modeling 278 11.3.3 Assessing agents’ value function for resource federation 279 11.4 An illustrative example 281 11.5 Conclusions 283 References 283 12 Fuzzy analytic hierarchy process using type-2 fuzzy sets: An application to warehouse location selection 285 I˙rem Uc¸al Sarı, Bas¸ar O¨ztays¸i, and Cengiz Kahraman 12.1 Introduction 285 12.2 Multicriteria selection 287 12.2.1 The ELECTRE method 289 12.2.2 PROMETHEE 289 12.2.3 TOPSIS 289 12.2.4 The weighted sum model method 290 12.2.5 Multi-attribute utility theory 290 12.2.6 Analytic hierarchy process 291 12.3 Literature review of fuzzy AHP 292 12.4 Buckley’s type-1 fuzzy AHP 293 12.5 Type-2 fuzzy sets 296 12.6 Type-2 fuzzy AHP 298 x CONTENTS 12.7 An application: Warehouse location selection 299 12.8 Conclusion 304 References 304 13 Applying genetic algorithms to optimize energy efficiency in buildings 309 Christina Diakaki and Evangelos Grigoroudis 13.1 Introduction 309 13.2 State-of-the-art review 312 13.3 An example case study 316 13.3.1 Basic principles and problem definition 316 13.3.2 Decision variables 318 13.3.3 Decision criteria 318 13.3.4 Decision model 320 13.4 Development and application of a genetic algorithm for the example case study 323 13.4.1 Development of the genetic algorithm 323 13.4.2 Application of the genetic algorithm, analysis of results and discussion 328 13.5 Conclusions 330 References 331 14 Nature-inspired intelligence for Pareto optimality analysis in portfolio optimization 335 Vassilios Vassiliadis and Georgios Dounias 14.1 Introduction 335 14.2 Literature review 336 14.3 Methodological issues 338 14.4 Pareto optimal sets in portfolio optimization 339 14.4.1 Pareto efficiency 339 14.4.2 Mathematical formulation of the portfolio optimization problem 340 14.5 Computational results 341 14.5.1 Experimental setup 341 14.5.2 Efficient frontier 342 14.6 Conclusion 344 References 345 Index 347 Preface Intherapidlyevolvingtechnologicalandbusinessenvironment,decisionmakingbecomes increasingly more complex from many perspectives. For instance, environmental and sustainable development issues have risen but the related policies, priorities, goals, and socio-economic tradeoffs are not well defined or understood in depth. Furthermore, tech- nological advances provide new capabilities in many areas such as telecommunications, web-basedtechnologies,transportationandlogistics,manufacturing,energymanagement, and worldwide trade. Finally, the economic turmoil increases the uncertainties in the global business environment, and has direct impact on all socio-economic and techno- logical policies. Such a challenging environment calls for the implementation of enhanced tools, pro- cesses, and techniques for decision analysis and support. Clearly, these should take into account the aforementioned multiple diverse aspects, in combination with the specific priorities and goals set by decision and policy makers. This has always been a prerequi- siteforprovidingdecisionsupportinarealisticcontext.Butitisnotenoughanymore,as decisionsupporttechnologiesshouldnowadaysalsoaccommodateavarietyofnew(often crucial) requirements, such as distributed decision making, the handling of massive and increasingly complex data and structures, as well as the computational difficulties that arise in building and using models and systems, which are realistic enough to represent the dynamic nature of existing problems and challenges posed by new ones. Multicriteria decision aid (MCDA) has evolved significantly over the past decades as a major discipline in operations research, dealing with providing decision support in complex, ill-structured problems involving multiple (conflicting) criteria. MCDA is involved in all aspects of the decision process, including problem structuring, model building, formulation of recommendations, implementation and support. Issues like pref- erence modeling, the construction of proper sets of criteria and their measurement, the characterization of different criteria aggregation models, the development of effective interactive solution techniques, and the implementation of sophisticated methods in user- friendlydecisionsupportsystems,havetraditionallybeenatthecoreofMCDAresearch. Among the many exciting trends developing in the area of MCDA, the one focused on exploring the connections of MCDA with other disciplines, is particularly interest- ing for providing integrated decision support in the complex context described above. In this framework, artificial intelligence (AI) has attracted much interest. Nowadays, AI is a broad field within which one can identify several major research areas, including

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.