Sensitivity Analysis: An Introduction for the Management Scientist PDF
Preview Sensitivity Analysis: An Introduction for the Management Scientist
International Series in Operations Research & Management Science Emanuele Borgonovo Sensitivity Analysis An Introduction for the Management Scientist International Series in Operations Research & Management Science Volume 251 Series Editor Camille C. Price Stephen F. Austin State University, TX, USA Associate Series Editor Joe Zhu Worcester Polytechnic Institute, MA, USA Founding Series Editor Frederick S. Hillier Stanford University, CA, USA More information about this series at http://www.springer.com/series/6161 Emanuele Borgonovo Sensitivity Analysis An Introduction for the Management Scientist 123 Emanuele Borgonovo Department ofDecision Sciences Bocconi University Milan Italy ISSN 0884-8289 ISSN 2214-7934 (electronic) International Series in OperationsResearch & ManagementScience ISBN978-3-319-52257-9 ISBN978-3-319-52259-3 (eBook) DOI 10.1007/978-3-319-52259-3 LibraryofCongressControlNumber:2017932423 ©SpringerInternationalPublishingAG2017 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpart of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilarmethodologynowknownorhereafterdeveloped. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt fromtherelevantprotectivelawsandregulationsandthereforefreeforgeneraluse. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authorsortheeditorsgiveawarranty,expressorimplied,withrespecttothematerialcontainedhereinor for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictionalclaimsinpublishedmapsandinstitutionalaffiliations. Printedonacid-freepaper ThisSpringerimprintispublishedbySpringerNature TheregisteredcompanyisSpringerInternationalPublishingAG Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland To Anna, Emma and Valeria, the joy of every moment of my life. Foreword Decision makers in a variety of diverse fields from engineering to finance are frequently aided by computer models that produce results that can be useful to decisionmaking.Thesemodelsareusuallyfairlycomplexsothequestionarisesas to what kinds of results would aid the decision maker. The computer, of course, alwaysproducesapointvalueoftheoutput;however,thisis(or,shouldbe)hardly satisfactory to the decision maker. At this point, it is useful to review how decisions are made when the options availablearesignificant,thatis,mayaffectthefinancialhealthofacompanyor,in the case of engineered systems, public health and safety. I had the opportunity to makesuchdecisionswhenIservedasacommissioneroftheUnitedStatesNuclear Regulatory Commission (2010–2014). A responsible decision maker (usually a group of managers) will rarely make a decision based on the point value that a computer program produces. More is usually required, such as the uncertainties about this point value, if the problem involves uncertain inputs, and the identifi- cation of the input parameters that contribute the most to the output and its uncertainties. The ultimate decision is essentially based on the judgment that is informed by this information and, possibly, by other considerations, such as applicable laws and regulations. This book contains a plethora of mathematical methods for extracting as much informationfrom the model aspossible. It isa rigorous treatmentthat isaddressed to the mathematically sophisticated reader. However, I believe that even the less sophisticated user will find it useful in that she can see what can be produced by modern methods and ask the analysts to exercise them so that a variety of insights can be obtained. You probably noticed that I used “she” in lieu of the more common “he” I did thisbecauseIfollowedProf.Borgonovo’srefreshingpracticethroughoutthebook. vii viii Foreword Therigorinthepresentationoftheanalyticalmethodsinthisbookisimpressive. This does not surprise me. Many years ago, Prof. Borgonovo did his Ph.D. thesis undermydirectionatMIT(hewasjustEmanuelethen).Iamgladthathisdesireto be rigorous has not diminished. George E. Apostolakis Professor Emeritus Massachusetts Institute of Technology Cambridge, USA Preface In late spring 2013, Prof. Camille Price invited me to write a book for the presti- giousOperationsResearchandManagementScienceSeries,publishedbySpringer and edited by Professor Frederick Hillier. I still clearly remember the intense meetingatthe2013INFORMSconferenceinMinneapolis,inwhichtheplanofthe book was discussed with professors Hillier and Price, and the Springer team, with MatthewAmboy,ChristineCrigler,andNeilLevine.Iheartilythankallofthemfor the support throughout the writing of this manuscript, which, as the reader may imagine, took longer than expected. Thismonographisintendedasanexpositoryintroductiontothemethodologyof sensitivity analysis of model output. It is primarily directed to students, investi- gators, and researchers that are familiar with mathematical models but are less familiar with the techniques for performing their sensitivity analysis. At the same time, the monograph does not aim to offer a review of all possible sensitivity analysis methods. A variety of sensitivity methods are available. Some of these methodsanswerquestionswhicharemodel-specific,whilesomeothermethodsare more general and apply to large families of models. The methods that we present here belong mainly to this second class of techniques. Nonetheless, the main goal ofthepaperistoraisetheawarenessoftheimportanceofsensitivityanalysis,asit plays a crucial role in the use of decision-support models. Performing sensitivity analysisrigorouslyallowsanalystsanddecisionmakerstoobtainanunderstanding of their model and insights from the model that would otherwise go lost. Iamextremelygratefulandmuchindebtwithallco-authorswithwhichIhave had the privilege to work during my career. My interest in sensitivity analysis initiates during my Ph.D. at MIT, with my advisor George E. Apostolakis. We started analyzing the important class of reliability importance measures and then moved on to global sensitivity analysis methods. Since then, I have had the priv- ilegeofworkingwithseveralco-authorsandfriends.Eachofthemhascontributed inbroadeningandenrichingmyviewofthemethodsandproblemsofmyresearch field. I will thank them in alphabetical order, hoping not to forget any of them: Hananeh Aliee, George Apostolakis, Manel Baucells, Francesca Beccacece, Valentina Bosetti, Greg Buzzard, William Castaings, Veronica Cappelli, Carmelo ix x Preface Cennamo, Alessandra Cillo, Francesco Di Maio, Marzio Galeotti, Stefano Gatti, Michael Glass, Gordon Hazen, Mary Hill, Kara Kockelman, Victor Jose, Marzio Marseguerra, Fabio Maccheroni, Massimo Marinacci, Giacomo Marangoni, Lorenzo Peccati, Alessandra Pellegri, Marco Percoco, Riccardo Polizzi, Elmar Plischke, Roberto Roson, Andrea Saltelli, Curtis Smith, Stefano Tarantola, Jurgen Teich,MassimoTavoni,FabioTonoli,RichardWendell,IrmelaZenter,EnricoZio, Stanley Zionts, and Maurizio Zollo. IwishtothankGeorgeE.Apostolakisforwritingtheforewordofthisbookand forinspiringsomuchofmyscientificlife.AspecialthankgoestoFabioTonoli,for his extreme dedication and hard work in carefully reading and spotting inconsis- tenciesinthemanuscript.IalsothankXuefeiLuandSumedaSiriwardenafortheir comments. At last, a truly special thank goes to my family, at large, and especially to my wonderfulwife,Anna,andmydearestdaughters,EmmaandValeria.Ithankthem heartily for the immense patience and constant support, especially when I was workinglatehoursorduringholidays.Theyarethejoyofeverymomentofmylife. Milan, Italy Emanuele Borgonovo
The list of books you might like
