volume 40 • issue 5 ims bulletinbulletin.imstat.org/wp-content/uploads/bulletin40_5.pdf · ims...
TRANSCRIPT
August 2011
Contents 1 Updated bulletin.imstat.org
website
2Members’ News: SteffenLauritzen;SimonTavaré,DonaldRubin,JonWellner,MarieDavidian,BaniMallick,DipakDey
4 Rick’s Ramblings: DehydratedElephants
5 IMS News: CouncilandExecutivemembers
6 Medallion Lectures: SylviaRichardson,Qi-ManShao
8 Anirban’s Angle: ANewCore?
9 Obituaries: TimothyRobertson,BernardHarris
11 ISOSS Meeting report
12 AOAS in the News
13 Terence’s Stuff:GenderEquity
14 IMS meetings
16 Other meetings
18 Employment Opportunities
20 International Calendar of Statistical Events
23 Information for Advertisers
Volume 40 • Issue 5
IMS BulletinBulletin blog site goes liveDimitrisPolitiswrites:Formanyyears,theIMS Bulletinhasbeenavailableintwoforms:ashardcopy,andasaPDFfilewhichwasanexactmirrorofthehardcopy(withaddedhyperlinks)availableasadownloadfromtheIMSBulletinwebsite,locatedathttp://bulletin.imstat.org.
InaccordancewiththeInstitute’saimtoencouragemembersto“gogreen”byoptingtoreceivetheelectronicversionoftheBulletin,andbepaperlesstotheextentpossible,theneedtorevamptheonlineBulletin’spresencewasbroughtforthbyPresident-ElectRuthWilliams.
WearenowinthehappypositiontoannouncetheinaugurationofthenewelectronicversionoftheBulletinthatwasmadepossibleviatheeffortsofourAssistantEditor,TatiHowell,andAndyStevens,whohasprovidedtechnicalsupport.The“new”Bulletin canbefoundatthesamewebsitehttp://bulletin.imstat.organd,aswellashavinganewlookandfeel,isbeautifullyinteractive.ThePDFversionwillstillbeavailableforeachissue,butnow,inaddition,selectedarticlesfromissueswillbeplacedonline,includingourregularcolumns,membernews,obituariesandmeetingreports.
Forthefirsttimereaderswillbeabletoposttheircommentsonanyarticletheyread,thusfacilitatingaquickandeasyexchangeofideas.Forexample,weareverymuchlookingforwardtoalivelydebatefollowingAnirbanDasGupta’scallfor“ANewCore’’forthestatisticsPhDcurriculum,onpage8ofthisissue,andonline.
Asever,werelyverymuchonIMSmemberstovolunteertheirnews;youcanstilldothisold-style,i.e.,byemailtobulletin@imstat.orgaswellasviathecontactpageoftheBulletinwebsite.
Wearealsopleasedtointroduceanewwebsitefeature,apermanentcolumnentitled“OpenForum”thatwillbeanopenblog-typeareaaspromisedintheJanuaryissue.Our“OpenForum”willbedevotedentirelytoreaders’viewsandcontributionsonmattersofcurrentimport.
WehopethatreaderswilltakeadvantageoftheseinteractivefeaturesofthenewonlineBulletin.Asastart,howaboutreadingsomearticlesandgivingussomefeedbackviathe“comment”feature?Welookforwardtohearingfromyou!
Executive CommitteePresident: PeterHall [email protected]: RuthWilliams [email protected] Past President: J.MichaelSteele [email protected]: JeanOpsomer [email protected] Secretary: GuentherWalther [email protected] Secretary:MartenWegkamp [email protected]
IMS Bulletin Editor: DimitrisPolitisAssistant Editor: TatiHowellContributing Editors: PeterBickel,AnirbanDasGupta,RickDurrett,NicoleLazar,TerrySpeed
To contact the IMS Bulletin: IMSBulletin
2LampernView,Uley,DursleyGL115TDUK
To contact the IMS regarding your dues, membership, subscriptions, orders or change of address:
IMSDuesandSubscriptionsOffice9650RockvillePike,SuiteL3503ABethesda,MD20814-3998USA
t 877-557-4674[toll-free in USA]t +12162955661[international]f +13016347099e [email protected]
To contact the IMS regarding any other matter, including advertising, copyright permission, offprint orders, copyright transfer, societal matters, meetings, fellows nominations and content of publications:
ExecutiveDirector,ElyseGustafsonIMSBusinessOfficePOBox22718,BeachwoodOH44122,USA
t 877-557-4674[toll-free in USA]t +12162955661[international]f +12162955661e [email protected]
2.IMsBulletin Volume40.Issue 5
IMS Members’ News
Contact information
IMSBulletinVolume40•Issue5August2011ISSN1544-1881
TheUK’sRoyalSocietyisafellowshipoftheworld’smosteminentscientistsandistheoldestscientificacademyincontinu-ousexistence.Fellowsareelectedforlifethroughapeerreviewprocessonthebasisofexcellenceinscience.Eachyear44newFellowsareelectedbyexistingFellows.AmongthoseelectedthisyeararetwoIMSFellows.
ProfessorSteffenLauritzenisaprofessorofstatisticsintheDepartmentofStatisticsattheUniversityofOxford,andaFellowofJesusCollege,Oxford.
Steffenisdistinguishedforhismanycontributionstothetheoryandpracticeofstatisti-calscience.Mostnotably,hepioneeredthefieldofgraphicalmodels,whichprovidesimpleandpowerfultoolsforencapsulatingconditionalindependencestructures.Theareaisnowacentralpartofmainstreamstatistics,andhasrevolutionizedmanyapplications.Lauritzen’scontributionsrangefromdeepinsightsintothemathematicalstructureofthisclassofmodels,throughinnovativeandefficientalgorithmsfortheiranalysis,toapplicationsinartificialintelligence,forensicscienceandmedicaldiagnosis.Otherimportantworkhascharacterizedtheanalyticstructureofstatisticalmodelsfromadifferentialgeometricperspective,viasufficiency
ProfessorSimonTavaréisaprofessorintheDepartmentofAppliedMathematicsandTheoreticalPhysicsandProfessorofCancerResearch,DepartmentofOncology,UniversityofCambridge,aswellasSeniorGroupLeaderintheCancerResearchUKCambridgeResearchInstitute.
Simonisaninternationallyrecognisedfigureintheinterfacebetweenstatistics,probabilityandbiologicalandmedicalsciences.Hehasmadeimportantcon-tributionstothestudyofcombinatorialstochasticprocesses,populationgeneticsandstatisticalbioinformatics.HisworkintheanalysisandinterpretationofDNAsequenceandrelatedgenomicdataincludesstatisticalaspectsofcoalescenttheory,includingthefirstfulllikelihood-basedmethodsfortheanalysisofsequencevariationdata,methodsforancestralinference,andforlikelihood-freeinferenceandapproximateBayesiancomputationforcomplexstochasticprocesses.Hehaspioneeredthedevelopmentofevolu-tionarygenomicapproachesforunderstandingcancer.
Royal Society elects Steffen Lauritzen and Simon Tavaré
Steffen Lauritzen
Simon Tavaré
IMS EditorsIMS Journals and PublicationsAnnals of Statistics:PeterBühlmannandTonyCai
http://imstat.org/aos Annals of Applied Statistics: BradleyEfron,Stephen
Fienberg,MichaelStein,KarenKafadar&SamuelKouhttp://imstat.org/aoas
Annals of Probability: OferZeitounihttp://imstat.org/aop
Annals of Applied Probability: AndrewBarbourhttp://imstat.org/aap
Statistical Science:JonWellnerhttp://imstat.org/sts
IMS Lecture Notes – Monograph Series http://imstat.org/publications/lecnotes.htm
IMS Collectionshttp://imstat.org/publications/ imscollections.htm
NSF-CBMS Regional Conference Series in Probability and Statistics:http://imstat.org/publications/nsf.htm
IMS Co-sponsored Journals and Publications
Electronic Journal of Statistics: DavidRupperthttp://imstat.org/ejs
Electronic Journal of Probability: BálintTóthhttp://www.math.washington.edu/~ejpecp
Electronic Communications in Probability:TimoSeppäläinenhttp://www.math.washington.edu/~ejpecp /ECP/index.php
Current Index to Statistics:GeorgeStyanhttp://www.statindex.org
Journal of Computational and Graphical Statistics:RichardLevinehttp://www.amstat.org/publications/jcgs
Statistics Surveys:LutzDümbgenhttp://imstat.org/ss
Probability Surveys:GeoffreyGrimmetthttp://imstat.org/ps
IMS-Supported Journals Annales de l'Institut Henri Poincaré (B): AliceGuionnet
http://imstat.org/aihpBayesian Analysis:HerbieLee
http://ba.stat.cmu.eduBernoulli:RichardDavis
http://isi.cbs.nl/bernoulliBrazilian Journal of Probability and Statistics:SilviaFerrari
http://imstat.org/bjps
IMS-Affiliated Journals ALEA: Latin American Journal of Probability and Statistics:
ClaudioLandimhttp://alea.impa.br/english
Probability and Mathematical Statistics:M.Musiela,J.Rosiński,W.Szczotka,A.Weron&W.A.Woyczyńskihttp://www.math.uni.wroc.pl/~pms
Other IMS contactsBulletin website:http://bulletin.imstat.org Main IMS website:http://imstat.org Managing Editor:[email protected] Production Editor:PatrickKelly
IMsBulletin . 3August . 2011
IMS Members’ News
Don Rubin awarded honorary doctorate by University of Bamberg, GermanyDonaldB.Rubin,JohnL.LoebProfessorofStatisticsatHarvardUniversity,hasbeenawardedanhonorarydoctoratebytheFacultyofSocialSciencesandEconomicsatOtto-Friedrich-UniversitätBamberg,Germany.Thehonorarydoctorateisinrecognitionofhisworkinthesocialsciencesandeconomics,andinparticular,forhisoutstandingcontributionstothefieldofappliedstatistics.
Jon Wellner received Humboldt Research AwardJonA.Wellner,UniversityofWashington,USA,hasbeenelectedtherecipientofaHumboldtResearchAward.HewasnominatedforthisprestigiousawardbyProf.TilmannGneiting,UniversityofHeidelberg,Germany.TheHumboldtResearchAwardisconferredinrecognitionoflifetimeachievementsinresearch.Inaddition,theawardeeisinvitedtocarryoutresearchprojectsofhisownchoiceincooperationwithspecialistcolleaguesinGermany,therebypro-motingfurtherinternationalscientificcooperation.Seehttp://www.humboldt-foundation.
de/web/programmes-by-target-group.htmlforfurtherinformation.
Dipak Dey associate dean of University of Connecticut College of Liberal Arts and SciencesDipakK.Dey,BoardofTrusteesDistinguishedProfessorattheDepartmentofStatistics,hassteppeddownasheadofthedepartmentandhasjoinedtheadministrationoftheCollegeofLiberalArtsandSciencesatUConnasassociatedean(PhysicalSciences).InadditiontohisdutiesoverseeingthePhysics,Chemistry,Mathematics,Geosciences,MarineSciencesandStatisticsDepartments,ProfessorDeywillalsobetakingcareofgraduateeducationwithintheCollegeofLiberalArts,internationalprogramdevelopment,andfosteringinterdisciplinaryresearchwithintheUniversity.
Marie Davidian elected ASA President-electMarieDavidian,NorthCarolinaStateUniversity,hasbeenvotedthenextPresident-electoftheAmericanStatisticalAssociation.ShewillserveasASAPresidentduring2012.
IMSFellowMarieDavidianisWilliamNealReynoldsProfessorofStatisticsatNCSU.SheisaFellowofASAandtheAmericanAssociationfortheAdvancementofScience,andanelectedmemberoftheInternationalStatisticalInstitute.ShehasservedaseditorofBiometrics andhasearnedseveralawards,includingtheAwardforOutstandingStatisticalApplication.
Distinguished Professor Bani Mallick BaniK.Mallickhasbeenpromotedtodistin-guishedprofessorinthedepartmentofsta-tisticsatTexasA&MUniversity.MallickisapioneerresearcherinthefieldofBayesiannonparametricregressionandclassifica-tionandtheauthorofBayesian Methods of Nonlinear Classification and Regression.Hedoesmajorcollaborativeresearchwithscientistsfromotherfieldssuchaspetro-leumengineering,bioinformatics,trafficmapping,superfundhazardouswastesites,andindustrialengineering.Hedevelopsnovelmethodologyandtheoryessentialforscientificresearchinthesecollaborationsandhassevenfundedresearchgrants.
4 .IMsBulletin Volume40. Issue 5
Isthereactuallyanyuseforadehydratedelephant?Rick Durrettthinksthatjustbecausewecansolveaproblem,itdoesn’tnecessarilymeanwe’reasking,oranswering,therightquestion…
Rick’s Ramblings: DehydratedElephants
Wellonmywaytobirthday“LX”(inan“XL”body),IamgratefulthatIhadtheopportunitytomeetsomeofthegreatestprobabilistsofourtime.MonroeDonsker,withhiswildandcrazyeyebrows,tookahiatusfromacademiabetweeninventinghisinvarianceprincipleandworkingwithVaradhanonlargedeviationsinordertoserveinthediplomaticcorps.OnenightatapartyinSpain,whilehearingabouttheexodusoftalentedindividualsfromthatcountryhequipped,“Ohyoumean,thedraininSpainismainlyonthebrain.”Thishastobeoneofthegreatestone-linersofalltime,evenifhewaslaterreprimandedforit.
Myseconddeardepartedhero,whogivesthetitletothispiece,isMarcKac,aself-described“simplepole.”Throughouthislife,herailedagainstdehydratedelephants.Naively,IthoughtthatintheageoftheInternetIcouldfindacopyoftheclassiccartoonwiththepunch-line:“I’mnotsurewhatyoucandowithadehydratedelephant,butitisnicetoseethatitcanbedone.”Inmyopinion,thisphraseappliestomanyofthearticlesthattheIMSpublishesitstwoprobabilityjournals.TheproblemwithourpublicationsisbestillustratedbyaclassicMarkKaclinethatcomplainsaboutpeoplewhostart with the answers and then figure out what is the question.Thejokebegins,“Theansweris:9w.”Afterasufficientlypregnantpause,thequestionisrevealed:“DoyouspellyournamewithaV,Mr.Wagner?”(“Nein, W”).Ifyouwanttolearnmoreaboutthemanbehindthesejokes,checkoutKac’sautobiography,Enigmas of Chance.
Returningtomydiatribe,oneoftheproblemsaboutwhatispublishedintheAnnals of Applied Probabilityliesinthedictumappliedbyreferees:“ifit’snot
hard,it’snotgood.”Itseemstomethatoneshouldgiveatleastequalweighttothequestion:doesthepapersaysomethinginterestingabouttheapplication?Toquotemyacademicgodfather,KaiLaiChung,fromtheprefaceofhisbookonMarkovChains:“Mathematiciansaremoreinclinedtobuildfirestationsthantoputoutfires.”Giventhecontentofourjournals,thequoteshouldbeupdatedto:“Oncewehaveablueprintforonefirestation,thereisnoneedtoactuallybuilditortoengageintheboringenterpriseofputtingoutfires.”
Thisisnottosaythattheoreticalprob-abilityisdeadandburied.ThepreviewofAliceGuionnet’sMedallionlectureonrandommatrices[in the previous issue]isveryexciting.AtmynewhomeinDurham,werecentlyhada2011revivaloftheSoutheasternProbabilityConference.Thefirstconferencebroughtanexcellentgroupofspeakers:TimoSeppäläinen(Wisconsin),JeremyQuastel(Toronto),LionelLevine(MIT),PeteKramer(RPI),DavarKhoshnevisan(Utah),SouravChatterjee(NYU),andGérardBenArous(NYU).AsIsatthereandwatchedthetalks,IcouldonlysayWOW.Ofcourse,muchoftheworkpresentedattheSEPChadamotiva-tioncomingfromphysicsorbiology,butsometalkslikeLionelLevine’slectureweremathematicallybeautifulsolutions.
Turning—asIinevitablydointhiscolumn—todiscussingmyownresearch,oneofthegreatthingsaboutcomingtoDukeayearagohasbeenthefactthatthe
medicalschoolisoncampus,ratherthan4+hoursawayfrom“centrallyisolated”Ithaca(anotherMarcKacphrasethatcouldeasilybeappliedtoSAMSI’slocationintheResearchTrianglePark).
Oneofmynewcollaborationswithlocalmedicalresearchersinvolvesastudyofovariancancer.In2010,anestimated21,800newcasesdevelopedintheUSwith13,850deaths.Asthenumbersmightsuggestthisisadeadlydiseasewitharapidtimecourse.Thefive-yearsurvivalrateis30%,dueprimarilytothefactthatthetumorswhichstartintheovaryandfallo-piantubesdonothavesignificantphysicalbarrierstometastasisintotheperitonealcavity.
WithaverytalentedDukeundergradu-ate,KavehDanesh,IhavebeenworkingtoapplyamechanisticmodelasanalternativetothephenomenologicalmodelsthathavewomenhoppingfromclinicalstageItoIItoIIItoIVaccordingtoaMarkovchain.Ourmodelsaremultitypebranchingprocessesinwhichtypeiindividualsgivebirthatrateai ,dieatratebi ,andmutatetotypei+1atrateui+1 .Beforeyouyawnandsaythatstudyingthesemodelsisnothardenoughtobeworthit,putdownthisissueoftheBulletin andtrytofigureoutwhattheasymptoticbehaviorofZ1hastodowithone-sidedstablelaws,anobservationthatprovidesusefulinsightsintotumorheterogeneityandyieldsremarkablysimpleexplicitformulasforvariousquantities.
[Continuesonnextpage]
August . 2011 IMsBulletin . 5
2011 IMS Election ResultsWearepleasedtoannouncetheresultsofthe2011IMSelections.ThenextPresident-ElectisHansR.Künsch.ThefivenewCouncilmembers,forathree-yearterm,are:SandrineDudoit,SteveEvans,SoniaPetrone,ChristianP.Robert,andQiweiYao.866people(21%ofmembers)participatedintheelections.Thanksforvoting!
Returningtothethemeofdehydratedelephants,inthisproject,KavehandIhavebenefitedfromemailexchangeswithVladisPipiras,SidResnick,andGennadySamorodnitsky,hydratedhumanswhohelpedusnavigatethevoluminousliteratureofstablelaws.LetΛbeaone-sidedstablelawwithindexα.WhytheheckPeterBrockwellandBruceBrowninanarticleintheZf Win1978decidedtolookatΛ–αandtoproduceapowerseriesforitsdensitythathastwogammafunctionsinthedenominator,Idon’tknow.Probably,theyjustwrotedowntheformulabecausetheyhadthetoolstodoso.However,ithasbeenusefulformakingconnectionswithKaplan-Meieresti-matesofthesizeoftheprimarytumorwhenthepatiententersthedeadlystageIIIofthedisease.
TowrapupthisRamble,thetake-homemessage,whichIlearnedalmosttwenty-fiveyearsagoinworkwithSimonLevinisthis:whenyourproblemscomefromrealquestions,thentheanswersleadyoutoaterritorythatismoreinterestingthanwhenyoustartwiththeanswerandfigureoutthequestion.
Grindingmyotherax:probabilitytheorywasdesignedtosolveproblems,butmuchofwhatispublishedislikepoetryinEsperanto,anesotericartformthatcanbeappreciatedonlybyahand-fulofpeople.Isthisreallywhatwewantinour“yellowjournalism”?
NEW IMS EXECUTIVE SECRETARYAuroreDelaiglewillbecometheIMS’sExecutiveSecretaryfromAugust2,2011toJuly9,2014,follow-ingMartenWegkamp’sthree-yearterm.
AuroreisaPrincipalResearchFellowintheDepartmentofMathematicsandStatisticsattheUniversityofMelbourne,withresearchinterestsinnonparametricestimationofcurves,measurementerrors,deconvolutionproblems,functionaldataanaly-sis,andhigh-dimensionalproblems.SheisanassociateeditorforJCGS,JRSSB,Statistica Sinica,theJournal of Nonparametric Statistics andtheJournal of the Korean Statistical Society.
Aurore’swebpageishttp://www.ms.unimelb.edu.au/~aurored/
TheExecutiveSecretary’sroleisexplainedintheIMSHandbook,theonlinever-sionofwhichisathttp://imstat.org/handbook/officers.html#ExecSec
HansKünsch
SoniaPetrone
SandrineDudoit
ChristianRobert
SteveEvans
QiweiYao
[Continuedfrompreviouspage]
6 .IMsBulletin Volume40. Issue 5
Theuseofregressionanalysestoprovideimportantinsightintoanarrayofproblemsisafundamentalapproachwithinthehealthsciences.Whetherthefocusisatthedetailedmolecularlevelorthebroadepidemiologicallevel,regressionanalysesaimtoteaseoutthelikelypartplayedbykeyvariables,ofteninthepresenceofcomplexsourcesofdiversityintheobservedandunderlyinghealthindica-tors.Theunderstandingofpatternsandsourcesofheterogeneityderivedfromthesemethodscanthenbeusedtoinformmorestructuredanalyses.
Fromastatisticalperspective,heterogeneityisnaturallymodelledviatheuseofarichclassofmodelsthatcanbroadlybereferredtoasmixturemodels.ThefundamentalstructuraluncertaintyassociatedwithsuchmodelsmakestheBayesianstatisti-calparadigmparticularlyattractivefordeliveringinterpretableinference.Bayesianmixturemodelshavebeenusedinthiswayinawiderangeofapplicationsspanningfrommolecularbiologyexperi-mentstodeterminedifferentialgeneexpression,togeographicalepidemiologyapplicationswhichstudydiseasepatternsacrossspacetosuggestpotentialenvironmentalriskfactors.Dependingonthecontext,interestcanbedirectedtowardsanumberoftasks,forexampletheinterpretationofsubgroupsortheclassificationandselectionofunusualpatterns.
Recently,attentionhasturnedtomultidimensionalmodelsfocussedonunderstandingthecombinedeffectsofmultiplepredic-tors(suchasbiomarkersorenvironmentalexposures)onthehealthofsubjects.Furthermore,manycurrentstudiesnowcollectdataondifferentstagesofbiologicalprocesseswiththeaimofenablingjointanalysisoflargesetsofpredictorsandofresponsesinordertodiscover,forexample,keypredictorsthatinfluencemanypheno-types.Thisisachallengingbi-directionalsituation,wherethenum-berofsubjectsisoftenmuchsmallerthanthenumberofpossiblepredictors,commonlyknownasthe‘largep,smalln’paradigm.
Insuchapplications,thesetofpredictorscanbeverylargeandalsocorrelatedinanintricateway.Thismeansthatinferenceviastandardregressionmodelsisplaguedbyinstabilityanddifficulty
ofinterpretation.Therearemanywaysinwhichprogresscanbemade,butthereducingthedimensionalityoftheproblemisoftenakeycomponentofsuccessfulanalyses.
Oneapproachtoaccomplishthisgoalistoincludewithintheanalysestheabilitytoclustermultivariatepatterns,usingthesubjectcovariateprofile,amultidimensionalvector,asthemainunitofinference.Adoptingaflexiblemixturemodel,thecovariateprofilesareiterativelyclusteredintogroups,whilstalsoallowingtheclusterassignmentstobeintroducedasrandomeffectsinaregressionmodeloftheresponse,therebypermittingtheresponsetoinfluencetheclusterassignment.Thus,itistheoveralleffectofcombinationsofcovariateswhichisassessedinasemi-parametricmanner.Althoughthetechnicaldetailsofsuchmodelsarefairlywellestablished,inordertoapplytheminepidemiology,itisimportanttofocusonissuesregardingtheinterpretabilityandexportabilityoftheinference.Carefulpost-processingoftheoutputisneededtocircumventlabellingissuesandtoallowthefulluncer-taintyoftheclusterassignmenttobepropagatedontotheinferenceforrandomeffectsofinterest.Asecond(andsometimescomple-mentary)approachfordimensionreductionistheinclusionofsomefunctionalitywithinthechosenmodeltoexplicitlyselectonlythosevariablesthatarestatisticallyassociatedwiththeresponse(s).Variableselectionforhighdimensionaldataisagenericcomponentofmanyanalyses.Inparticular,ingeneticepidemiologyoneofthemostsought-afterstatisticaloutputisasetofreproducible‘sparse’modelsthatselectonlyafewrelevantpredictors(i.e.SNPs,transcripts)amongstaverylargesetofpossiblecandidates,togetherwithgoodassessmentofhowuncertaintheirroleis.Thequestionofhowtobuildsuitablepriormodelsthatachieveacompromisebetweentheaimsofcontrollingsparsityandthatofborrowinginformationacrosstheresponsesinordertoenhancediscoveryofkeypredictorsisanactiveareaofresearch.
Iwilldiscusssomerecentresultsintheseareas,togetherwithcomputationalissues.Thetalkwillbeillustratedthroughoutbycasestudiesfromepidemiologyandintegrativegenomics.
SylviaRichardsonhasheldtheChairofBiostatisticsintheDepartmentofEpidemiologyandBiostatisticsat
ImperialCollegeLondonsince2000,andheadstheBiostatisticsgroup.AfterafirstdegreeinMathematics,she
receivedaPhDinprobabilitytheoryfromtheUniversityofNottinghamandaDoctorat ès sciences fromUniversité
ParisXI.SheheldlectureshippositionsatWarwickUniversityandUniversitéParisVandsubsequentlybecame
“DirecteurdeRecherches”intheFrenchResearchInstituteforMedicalResearch.ShewasawardedtheGuyMedal
inSilverfromtheRoyalStatisticalSocietyin2009andholdsaRoyalSocietyWolfsonResearchMeritawardforher
workonBayesianintegrativeanalysisinepidemiology,systemsbiologyandgenomics.HerMedallionLectureis
atJSMMiami,onTuesday,August2,at10:30am.
Medallion Lecture preview: SylviaRichardson
Recent developments in Bayesian methods for discovering regression structures: applications in the health sciences
August . 2011 IMsBulletin . 7
Qi-ManShaoisChairProfessorintheDepartmentofMathematicsattheHongKongUniversityofScienceand
Technology.Hereceivedabachelor’sdegreeinMathematicsandamaster’sdegreeinprobability/statistics
fromHangzhouUniversityandthenreceivedaPhDinprobability/statisticsfromtheUniversityofScienceand
TechnologyofChina,undertheguidanceofProfessorXiruChen.HeisaFellowofIMS,andwasaninvited
speakerattheInternationalCongressofMathematiciansin2010.
Qi-Man’sprimaryresearchinterestsincludelimittheoryinprobability,largesampletheoryinstatistics,
self-normalizedlargedeviationtheory,high-dimensionalandlarge-scalestatisticalanalysis,Stein’smethodfor
normalandnon-normalapproximation;Gaussianrandomfields,randompolynomialsandmatricesandMonteCarlostudies.Hehas
servedontheeditorialboardsofAnnals of Applied Probability andAnnals of Statistics.
Qi-ManwillgivehisMedallionLectureatJSMMiami,onJuly31at4:00pm.
Medallion Lecture preview: Qi-ManShao
Normal and Non-normal Approximation by Stein’s MethodLetWnbeasequenceofrandomvariablesofinterest.AssumethatWnisasymptoticallystandardnormal.Whentheaccuracyofthenormalapproximationisthecentralconcern,therearemainlytwoapproachesforestimatingtheerroroftheapproximation.Oneapproachistostudytheabsoluteerrorsupz|P(Wn≥z)−(1−Φ(z))|viaaBerry-Esseentypebound.TheotherapproachistoestimatetherelativeerrorofP(W≥z)to1−Φ(z)throughaCramértypemoderatedeviation.AstandardandclassicalapproachtoBerry-EsseentypeboundandCramértypemoderatedeviationisFouriertransformand/orconjugatemethods.
However,therearemanysituationswhere,forexample,forWnunderdependencestructure,theaforementionedmethodsareverydifficulttouse.ItisStein(1972)whointroducedatotallydifferentmethodtodeterminetheaccuracyofthenormalapproximationtothedistributionofasumofdependentrandomvariables.Themethodhasbeenshowntobeapowerfultoolinestimatingaccuracyofvariousprobabilityapproximations.Incontrasttotheclassicalcharacteristicfunctionmethod,themeritofStein’smethodisthatitworkswellforbothindependentanddependentrandom
variables.Themethodhasbeenextendedbeyondthenormalapproximationandappliedtoproblemsinvariousareas.
Inthislecture,wewillgiveabriefsurveyofrecentdevelop-mentsofStein’smethodforbothnormalandnon-normalapproximation.Motivatedbyapplicationstofalsediscoveryrateinsimultaneoushypothesistests,ourfocuswillbeontheCramér-typemoderatedeviationsforStudentizednon-linearstatistics,andforWnunderageneralframeworkoftheSteinidentitywhichincludesthesumofrandompermutationarrays,thebinaryexpansionofarandominteger,theanti-votermodelonacompletegraphandtheCurie-Weissmodelasspecialcases.Ontheotherhand,whenthelimitingdistributionofWnisunknown,anexchangeablepairsapproachwillbeintroducedtoidentifythelimitingdistributionandevaluatetheabsoluteerrorofapproximation.ApplicationstoCurie-Weissmodelatthecriticaltemperatureandothermodelsinstatisticalphysicswillalsobediscussed.
ThereseemtobemanyunsettledquestionsandmuchterritorythatcanbeexploredwiththehelpofStein’smethod.Indeed,thepotentialofStein’smethodisunbounded.
IMS activities at JSMIfyou’recomingtotheJointStatisticalMeetingsinMiami,Florida(July30toAugust4),pleasedropbytheIMSboothintheExpotosayhello.Ifyourstudentswillbethere,remindthemthatIMSmembershipisfree,andprovidesthemwithonlineaccesstoalltheIMSjournals—theycanjoinupattheIMSboothoronline.
TherewillbeanopportunitytosocializewithotherIMSmembersandJSMattend-eesaftertheIMsPresidentialAddressandAwardsCeremony,onMondayAugust1from8:00–11:00pm,inBallroomDattheConventionCenter.Hopetoseeyouthere!
ForprogramandotherinformationaboutMiami,pleaseseetheJSMwebsite,athttp://amstat.org/meetings/jsm/2011/
8 .IMsBulletin Volume40. Issue 5
AnirbanDasGuptawrites:PhDstudentsinstatisticsdepartmentsacrosstheworldareaskedtotakeacourseonthecoretheoryofinference,theso-called“qualifiertheorycourse”.ItookmineattheISIinCalcuttain1977.ItwasmasterfullytaughtbyK.K.Roy,andcoveredwhatwasessentiallygloballyregardedasthecoreofinference,exponentialfamilies,sufficiency,ancillarity,complete-nessandUMVU,MLEs,FisherinformationandCramér-Rao,asymptoticsofMLE,consistency,deltaandSlutsky’stheorem,NPlemma,MPandUMPtests,MLRfamilies,LRT,UMAconfidencesets,dualityofconfidenceandtesting,basicgametheory,Bayesrules,admissibility,minimaxity,Wald’sSPRT,andranktests.BickelandDoksumhadjustcomeoutandhadn’treachedIndia;weusedFerguson(1967),andweallagreedthatitwasafulfillingcourse.
Butthatwasthen,andthisisnow.Throughpurelypersonalinteractions,Ihavehadanecdotalevidenceofsomesentimentthatapartofwhatwaslongregardedasthecoreofinferenceisnotconsideredtoorelevantnow.Theeventsandthediscoveriesofthelastthirtyyearsmakeitnecessarytore-evaluatewhatisthecoretheoryofstatisticsthatafreshPhDoughttobeexpectedtoknow,andunderstand.Togetasenseofmycolleagues’pulse,IcheckedoutthesyllabiofthecorecourseatBerkeley,Stanford,Chicago,Washington,UPenn,CarnegieMellon,andDuke.Ialsocontacted11expertsintheUS,Europe,Australia,andIndia,andsolicitedtheirdefinitionofwhatshouldbeinthequalifiertheorycourse.Theresponsesquitesurprisedme.
Iwassurprisedbythefantasticdiversityoftheopinionsonwhatshouldbeinthatfirsttheorycourse.Theintersectionofthedefinitionswasempty,barringsufficiency,exponentialfamilies,andMLEs.Buttherewasanunmistakabledesiretoputlessemphasisonparametrics,unbiasedness,onUMPtests,oncertainpartsofdecisiontheory,traditionalsequentialanalysis,andtheoldnonparametrics.Ontheotherhand,theresponsesincludedmany“newage”topics:thebootstrap,AICandBIC,VCtheory,empiri-calprocesses,permutationtests,EM,MCMC,functionestimation,sparsity,causalinference,extremevalues,andsomemore.
Theresponsesarerevealing.Theytoldmethatwhilethereisasharphungerforchangeinthattraditionalcorecourse,itisnolongerpossibletohaveanapproximateglobalconsensusonwhatthatfirstcourseshouldteach.Thecorecoursewillprobablybecomeratherlocal,andwewouldn’tbeabletoassumethatafreshPhDfromastatisticsprogramhasseenandbeentestedonasetofcommontopicsininference.
WhatwouldIteachpersonallyinthatfirstcourse?AfterreadingalltheresponsesthatIreceived,IthinkIpersonallyagreewiththesetwostatementsfromJohnMardenandPhilipStark:the first math-stat course should be about how to think about models and infer-ence and the mathematical (as opposed to computational) framework to attack the problems;and,you need to have some idea of what’s possible, and where to look for approaches, ideas, inspirations, theorems.Ithinkmyowndottrina à nouveau couldbesomethinglikethis,thetopicsandthenumberof50-minutelecturesoneach(43intotal).
Problemsandbasicprinciplesofinference,selectingandevaluatingaprocedure,lossandrisk,bias,variance,parametricvs.nonparametricmodelling,optimalityvs.robustness,Hogg’sadaptiveestimate(nontechnical;2);modelling,location-scale,expo-nentialfamilies,mixtures,heavytails,non/semi-parametricmodels,dependence(4);datasummary,likelihoodfunction,sufficiency,factorizationandRao-Blackwell,definitionofUMVU(3);scorefunction,Fisherinformation,informationmatrix,Cramér-Rao(3);MLEs,generalandinexponentialfamily,somenonregular,somemultiparameter,nonexistence,difficultyofcomputing,refertoEMinBickel-Doksum(3);simulatetheCauchyMLEandmedian,statementofPortmanteau,asymptoticnormalityandCramér-Raoconditions,observedinformationandsandwich,twoparameterGamma,plug-in,deltaandSlutsky,applications(5);priors,poste-riors,conjugatepriors,posteriormeans,posteriorandBayesrisk,comparisonwithMLE,Bayesvs.minimaxity,fromBickel-Doksum(5);testing,errorprobabilitiesandpower,NPlemma,applica-tions,statementofUMPonesidedtests(3);LRT,threeexamplesofBickel-Doksum,chisquarelimit(3);confidencesets,dualitywithtesting,t interval,asymptoticconfidenceintervals,defini-tionofposteriorcredibleinterval(3);theempiricalCDF;SLLN,Glivenko-CantelliandDKW;purpose,use,andscopeofbootstrap;bootstrapbiasandvarianceestimation;consistencyofbootstrap,permutationtests(4);apersonalselectionfromJames-SteinandDonoho-Johnstoneestimates,sparsity,SURE,Gaussiansamplemaximum,kernels,RKHSandfunctionestimation,modelchoice,AIC,BIC,VCandmartingaleinequalities,Bayesfactors,Dirichlet
process,Bernstein-vonMises(5).Perhapssomeoneelsewilladdresswhatshould
betaughtinStat100.Opinionswillprobablydif-feronthat,too.Butthat’sanissueforanotherday.
Anirban’s Angle: ANewCore?
LogontotheBulletinwebsiteandshare
yourcommentsathttp://bulletin.imstat.org
August . 2011 IMsBulletin . 9
ProfessorTimRobertsonpassedawayonApril5,2010,totheregretofhiscol-leagues,familyandfriends.Hisformerstudents,collaboratorsandcolleagueswhohadheldhiminhighesteemthroughouthisdistinguishedacademiccareercovering39yearsintheDepartmentofStatisticsandActuarialScienceatTheUniversityofIowamournedhispassing.
ProfessorRobertsondiedsuddenlyattheageof72.HewasprecededindeathbyhislovingwifeandlifelongcompanionJoan,whohadpassedawayonFebruary25,2010.Heissurvivedbyhisfourchildren--Kelly,Jana,DougandMike--andsixgrandchildren.
TimothyJoelRobertson,sonofHelenOliver-GirdnerandstepfatherFlickGirdner,wasbornOctober4,1937,inDenver,CO.Tim’sfamilymovedtoChilicothe,MO,in1942.HegraduatedfromChilicotheHighSchoolin1955.TimattendedtheUniversityofMissouriinColumbiaandreceivedhisB.A.inmath-ematicsin1959.HeandJoanweremarriedpriortohissenioryearatM.U.TimenteredGraduateSchoolinthefallof1959,andobtainedhisM.S.inmathematicsin1961.TimacceptedanAssistantProfessorpositioninmathematicsatCornellCollegeinMountVernon,IA,andheandJoanalsoservedashouseparentsinthemen’sdormi-toryfrom1961–1963.
TimandJoanenjoyedtheenvironmentatCornellCollege,buttheydecidedtoreturntoColumbia,MO,in1963forTimtopursuehisPh.D.Tim’sthesisadvisoratM.U.wasH.D.(Dan)Brunk,andheandTimestablishedaclosefriendshipwhichexisteduntilDan’sdeathin2009.DaninstilledanappreciationformathematicalrigorandathrillofresearchinTimthatcarriedthroughTim’sentireacademiclife.
TimjoinedtheDepartmentofStatisticsandActuarialScienceattheUniversityofIowaasanewPh.D.in1965.Timwasagoodteacherandresearcher,andwaspromotedtoAssociateProfessorin1968,andtofullProfessorin1974.EighteenstudentswrotetheirPh.D.dissertationsunderProfessorRobertson’sguidance.Mostbecamesuccessfulscholarsintheirownrights,andseveralmaintainedcontactwithProfessorRobertsonuntilhisdeath.
ProfessorRobertsonwasasuccessfulandwell-respectedscholar.HewasaFellowofboththeAmericanStatisticalAssociationandtheInstituteofMathematicalStatistics.HewaselectedtomembershipintheInternationalStatisticalInstitute,andhewasanactivememberoftheMathematicalAssociationofAmerica.ProfessorRobertsonwastheleadauthor(withF.T.WrightandR.L.Dykstra)onthemonograph“OrderRestrictedStatisticalInference,”whichwasconsideredthestandardreferenceonthetopic.TimwastherecipientofaUniversityofIowaCollegiateTeachingAwardin1990andwasgenerallyrecognizedasbeinganoutstandingteacher.TimservedasAssociateEditorforseveralstatisticaljournals,andalsotookanactiveroleinUniversityactiv-ity.HeservedontheEducationalPolicyCommitteeforseveralyearsaswellasontheBoardinControlofAthletics.ThoughTimwasademandingadvisor,hewasalsonurturing.Thoughhehadhighexpectationsofhisstudents,hisofficedoorwasalwaysopentothem.Timwasenthusiasticintheclassroom,andmanyofhisstudentsrecallbeingkeptontheirtoesbyrelevantques-tionsbeingdirectedatthem.
TimeventuallydevelopedmanyinterestsoutsidetheUniversity.Timlovedtheoutdoorsandoftentookhisfamilyoncampingandcanoeingtripstothe
Obituary: Timothy Robertson1937–2010
Tim Robertson
BoundaryWatersofMinnesotaandvariousIowarivers.Timtrainedhuntingdogs,competedinfieldtrialswiththem,andwasafoundingmemberoftheEasternIowaShootingDogAssociation.In1992,TimandJoanpurchasedafarmneartheCedarRiveratCedarBluffs.Theybuiltamodesthome(referredtoastheCabin)onthefarm,andtheCabinwaslaterenlargedtobecometheelegant,comfortablehomethatheandJoanenjoyeduntiltheirdeaths.
Timlovedhischildren,grandchildren,children’sspouses,dogs,jawbreakers,horses,fishing,redsweatshirts,beautifulsunsets,risquésongs,hisfourwheeler,ridinghisbiketowork,agooddebate,andpopcorn.Timwasamanofgreatintellect,passion,laughter,strengthandlovethatwillbemissedbyallwhoknewhim.
Richard DykstraProfessor of Statistics & Actuarial Science
College of Liberal Arts and SciencesThe University of Iowa.
This obituary first appeared on the University of Iowa’s website and is reproduced
by kind permission,
10 .IMsBulletin Volume40. Issue 5
BernardHarrispassedawaypeacefullyonJanuary28,2011,atTulaneUniversityMedicalCenterinNewOrleans,La.,fol-lowingcomplicationsfromheartsurgery.
BernardwasbornonJune20,1926,inNewYorkCity,thesonofSamuelS.andEllaL.(Heyman)Harris.Anacademicallyprecociousyoungster,hegraduatedatanearlyagefromTownsend-HarrisHighSchool,andthenenteredCityCollegeofNewYork.Duringhiscollegeeducation,hewasdraftedintothearmy,assignedtotheCounterIntelligenceCorps,andsenttoGermanytowardstheendofWorldWarII.Uponcompletionofmilitaryservice,hefinishedhisbachelor’sdegreeinbusinessadministrationatCityCollegein1946.Hechangedhisacademicfocustomathematicsandstatistics,earnedamaster’sdegreefromGeorgeWashingtonUniversityin1953,andinoneyear,completedhisdoctorateatStanfordUniversityin1958.
Duringtheyearsbetweenhismaster’sanddoctoratedegrees,heworkedasastatisticianattheUSCensusBureauandasamathematicianfortheNationalSecurityAgency.Aftercompletinghisdoctorate,hebecameanassociateprofessorintheDepartmentofMathematicsattheUniversityofNebraska–Lincoln.HemovedtoMadison,Wisconsin,toworkasaprofes-sorintheMathematicsResearchCenterattheUniversityofWisconsin–Madisonfrom1964to1985,andasaprofessorintheStatisticsDepartmentattheUniversityofWisconsin-Madisonfrom1966to2002.AfterhisretirementfromUW–Madison,herejoinedthefacultyoftheStatisticsDepartmentattheUniversityofNebraska-Lincolnandwasanadjunctfacultymemberthereuntilhisdeath.Bernardenjoyedvisit-ingprofessorshipsattheTechnicalInstituteinMunich,Germany;TechnicalUniversity
ofEindhoven,Netherlands;UniversityofLund,Sweden;theMathematicsInstituteSteklova,Moscow,Russia;UniversityofMünster,Germany;HeinrichHeineUniversityinDüsseldorf,Germany;andKTHinStockholm.
BernardwasamemberofmanycommissionsandadvisoryboardsforthegovernmentincludingareviewboardoftheUSNuclearRegulatoryCommissionandtheStatisticsTaskForcefortheFAA/DODCommitteeonMaterialProperties.HewasanelectedfellowoftheInstituteofMathematicalStatisticsandtheAmericanStatisticalAssociation.HewasproudtobeoneoffoundingmembersoftheSectiononRiskAnalysisfortheAmericanStatisticalAssociationandservedasitsfirstchair.HewasamemberoftheClassificationSocietyofNorthAmerica,andservedonitsboardofdirectors.HewasalsoamemberoftheInternationalClassificationSocietyandtheAmericanMathematicalAssociation.HewasarecipientofthePioneersofScienceAward.ProfessorHarriswasaperennialadvocateofandcontributortostatisticalsciencefortheDepartmentoftheArmy.HeparticipatedaspartoftheMathematicsResearchCenteratWisconsinthatsup-portedtheArmyinaddressingresearchquestionsandpresentedhisworkatcount-lessstatisticalconferencessponsoredbytheArmyspanningtheyears,1964–2010.
Hiscontributionstoriskanalysis,reli-ability,probability,andstatisticalinferenceswithapplicationtoopenDODquestions,suchasthesurvivabilityofsubterraneantargets,wasrecognizedin1982withtheWilksAwardforContributionstoStatisticalMethodologiesinArmyResearch,DevelopmentandTesting.Hisworkcontinuedtoaddresscurrentproblemsinhislateryears,withrecentworkconcerning
Obituary: Bernard Harris1926–2011
mathematicalmethodsincombatingterrorism,andhis2010paperentitled“RandomContaminationofSemiconductorMaterials.”
Bernardwastheauthorofabook,Theory of Probability, andtheeditorofSpectral Analysis of Time series andGraph Theory and its Applications.Hepublishedhundredsofarticlesandreviewsinprofes-sionaljournalsoverthecourseofhiscareer.Hewasmostproudofhisworkinrandommappings,combinatorics,reliabilityandriskanalysis.
Inadditiontohisacademicinterests,Bernardenjoyedawidevarietyofmusic(classical,operaandjazz),readinginseverallanguages,gourmetcooking,filmsanddoingtheNew York Timescrosswordpuzzles(inink).Helovedtocreatesimpleandcomplexpuns.
Mostofall,Bernarddeeplylovedandappreciatedhisfamilyandsteadfastlysup-portedthemwithhisadviceandpresence.BernardwasamemberoftheBethIsraelCongregationinMadison.
BernardmarriedAnita(Greenberg)in1949.HeandAnitahadfourchildren,ShelleyNolte,MarkHarris(Katherine),DavidHarris(Susan)andSusanHandrich(Tom).Hiswife,Anita,diedin1977.HemarriedSusanStephensBurnsin1983,andwasastepfathertoLauraBurns(BradSinner)andErinCharles(Jesse).Hewastheproudgrandfatherofseven.Hewaspre-cededindeathbyhisparents;hisbrothers,AlanandAlvin;hisfirstwife,Anita;andhisbeloveddaughter,Shelley.
MemorialsmaybedirectedtotheAmericanStatisticalAssociationtosupporttheeducationofyoungstatisticians.
Written by members of the Harris family. Reproduced with permission from
http://madison.com/obit/179112
August . 2011 IMsBulletin . 11
Meeting report: ISOSS Lahore Conference
MunirAhmed,theFoundingPresidentofIslamicCountries
SocietyofStatisticalSciences(ISOSS),reportsontheLahoreISOSS
Conference,whichemphasizedtheincreasedstrategicroleof
statisticians.
The8thInternationalConferenceonRecentAdvancesinStatistics:“Statistics,BiostatisticsandEconometrics,”washeldonFebruary8–9,2011attheNationalCollegeofBusinessAdministrationandEconomics(NCBA&E),Lahore,Pakistan.TheconferencewasorganizedbytheIslamicCountriesSocietyofStatisticalSciences(ISOSS)incollaborationwiththeNCBA&EinhonouroftheISOSSPresident,DrShahjahanKhan,forhisoutstandingcontributioninleadingISOSSasaninternationalprofessionalorganisationofstatisticiansthroughorganisationofanumberofinternationalconferencesandpromotingstatisticsglobally.
Delegatesfromaroundtheworld,includingAustralia,Europe,MiddleEast,SouthAsiaandUSA,presentedtheirrecentresearchoutputsindiverseareasofstatistics.Atotalof320participantsfromPakistanandothercountriestookpartintheconferencedeliberations.Invariousscientificsessions,142researchpaperswerepresentedhighlightingrecentadvancesinstatistics,biostatisticsandeconometrics.Unfortunately,manyWesternparticipantswereunabletopresenttheirsubmittedpapersduetotravelrestrictionstoPakistan.AlargenumberofstatisticianswhocouldnotjointheconferencesenttheirmessagesofappreciationemphasizingthegreatcontributionsofDrShahjahanKhan.
MianShamimHaider,aformerFederalMinisterofRailwaysandSportsoftheGovernmentofPakistan,inauguratedtheconfer-enceasthechiefguest.Heemphasizedtheuseofstatisticsinplan-ninganddevelopmentofPakistanandhowhigh-techindustriescanbeestablishedinPakistanbyimportingthelateststatisticaltechniquesanddevelopingnewindigenoustechnologiesbasedonrecentadvancesinPakistan.ThefoundingPresidentofISOSSDrMunirAhmad,andVice-PresidentofISOSSDrMohammadHanifMian,spokeintheopeningsession.AllthespeakersspokehighlyofthegroundbreakingworkofDrShahjahanKhan,andthechiefguestpresentedacresttohimtomarkhisremarkablecontributionstothepromotionanddevelopmentofstatistics.
InhisaddressDrShahjahanKhanhighlightedhowthepercep-tionofstatisticshaschangedfromameretoolofcalculationofnumberstomakingdecisionsinthefaceofuncertainlyandtakingtheroleofstrategistsforgovernmentandbusiness.Heemphasizedmorevisibleanddominantpresenceofstatisticstothesolutionofcomplexissuesthehumanityfacestoday.Statisticiansareuniquelypositioned,hesaid,tocontributesignificantlyinaddressingthe
burningproblemsofourtime.Statisticshascomeupwithsolutionstomanyproblemswhereotherdisciplineshavefailed.
OneofthemainattractionsoftheconferencewastheofficialopeningofISOSSHouse,whichhadbeenrecentlycompleted.TheparticipantsreflectedonthedreamofthefoundingPresidentofISOSS,DrMunirAhmad,andthankedeveryonewhohadcontributedtotheprojectfinanciallyandotherwise.TheofficeofISOSShasnowbeenshiftedtotheISOSSHouseandithasfacili-tiesforoffices,library,computerlab,trainingandmeetingroomsetc.Inthenextphaseoftheproject,ISOSSplanstobuildfacilitiestoaccommodatevisitingscientistsandresearchworkerstopromoteexcellenceinstatisticalteaching,learningandcollaborativeresearch.
TheBusinessSessionoftheconferencerecommendedunani-mouslythat:inPakistantheNationalStatisticalCouncilmustberevived,therecently-createdPakistanBureauofStatisticsshouldbeproperlystaffedbyqualifiedstatisticalscientists,Pakistanshouldtakenecessarystepstomeettherequirementsforbecomingastatistically-advancedcountry,theGovernmentmustworkoninternationalbodiesparticularlyintheUNO,UNDPetc.forinvit-inglocalstatisticalexpertsbyothercountriesasadvisors,establishStatisticalResearchStationineachprovinceforthediscoveryofnewknowledgeanddevelopmentofnewtechniques,conductmicro-censusesateveryalternateyearfortheupdatingofpopula-tiongrowth,andaddressanimmediateneedforanInternationalJournalofOfficialStatistics.
Dr Munir Ahmed (left), former Minister Mian Shamim Haidar presenting the crest of honour to Dr Shahjahan Khan, University of Southern Queensland, Australia
12 .IMsBulletin Volume40. Issue 5
Ann Appl Stat paper in “Hockey Stick” debateThe “Hockey Stick” Debate: new study finds “substantial uncertainty” with temperature reconstructions using natural proxiesApressreleasefromtheKelloggSchoolof
Management,homeofoneoftheauthors
ofacontroversialpaperintheAnnalsof
AppliedStatistics,outlinesafewofthe
issues:
Since1998,climatescientistshaveattemptedtoreconstructglobalannualtemperatureoverthelastmillenniumusingnaturalproxiessuchastreeringsandicecores.However,anewstudyfindssubstan-tialuncertaintyinthesereconstructions.
Intheirresearch,[IMSmembers]ProfessorsBlakeleyMcShaneoftheKelloggSchoolofManagementatNorthwesternUniversityandAbrahamWyneroftheWhartonSchoolattheUniversityofPennsylvania,findthatnaturalproxiesareweakpredictorsofglobalannualtempera-turesoverthelastonethousandyears.
Proxy-basedtemperaturereconstructionssuchastherenowned“hockeystick”graphhavearguablybecomethemosticonicillustrationofglobalwarming.Thehockey
stickisatemperaturereconstructionwhichfeaturesalongflat“handle”indicatingtem-peratureswererelativelyconsistentfor900yearsfollowedbyasharpupward“blade”indicatingthattemperaturesdramaticallyincreasedoverthepast100years.
“Weconcludeunequivocallythattheevidencefora‘long-handled’hockeystickislackinginthedata,”theauthorswrote.Thehandleofthehockeystick“isbestunder-stoodtobeafeatureofregressionandlessareflectionofourknowledgeofthetruth...Thefundamentalproblemisthatthereisalimitedamountofproxydatawhichdatesbackto1000AD;whatisavailableisonlyweaklypredictiveofglobalannualtempera-ture.”Whiletheydonotdenytheexistenceofclimatechange,theybelievethecaseasitpertainstoproxy-basedtemperaturereconstructionshasbeenoverstated.
Intheirstudy,thestatisticiansassessedthereliabilityoftemperaturereconstruc-tionsandtheirstatisticalsignificanceagainstvariousmodels.Additionally,theycomparedexistingreconstructionsofNorthernHemisphereaverageannual
landtemperatureoverthelastmillenniumtotheirownreconstruction.Whiletheirreconstructionproducedtemperatureesti-matesoverthelast500to1000yearswhichlargelymatchedthoseproducedbyclimatescientists’models,McShaneandWynerfoundsubstantiallylargeruncertaintyinter-valsascomparedtotheothermodels.
“Naturalclimatevariabilityisnotwellunderstoodandisprobablyquitelarge,”theauthorswrote.“Itisnotclearthattheproxiescurrentlyusedtopredicttempera-tureareevenpredictiveofitattheannualordecadalscaleoverseveralcenturies.Nonetheless,paleoclimatologicalreconstruc-tionsconstituteonlyonesourceofevidenceintheAnthropogenicGlobalWarning(AGW)debate.”
Thestudy,“AStatisticalAnalysisofMultipleTemperatureProxies:AreReconstructionsofSurfaceTemperaturesOvertheLast1000YearsReliable?”waspublishedinMarch2011issueofThe Annals of Applied Statistics.
AOAS Editor’s commentBradleyEfron,Editor-in-ChiefofThe Annals of Applied Statistics,notesthatthearticlewaspublishedwith13commentar-ies,bybothstatisticiansandclimatespecial-ists,mostofwhichwerequitecritical.
Oh? Not that kind of hockey stick?
In Other News: AOAS in the News!TheNew York Times featuredanarticleonJuly8aboutfalsecancercuremedicine.Thearticle,byGinaKolata,creditsThe Annals of Applied Statisticswithpublishingthekeywhistle-blowingarticle:apaperbyKeithBaggerlyandKevinCoombes(AOAS2009,Vol.3,No.4).ReadKolata’sarticleathttp://www.nytimes.com/2011/07/08/health/
research/08genes.html
August . 2011 IMsBulletin . 13
InFebruary2010thedirectorofmy(medicalresearch)institutecreatedacommitteetoformulatepoliciesand
actionspromotinggenderequity(GE),withtheunderstandingthathewoulddowhateverwasinhispowertoimplementitsrecommendations.
Whatistheproblem?Statisticscanhelpusseeit.Trendsinthepercentageoffemalesatdifferentacademiclevelscanbeaneye-opener.Inmyinstitute,nearly60%oftheundergraduateandPhDstudentsarewomen,abouthalfofourpost-doctoralscientistsarewomen,25%ofourlabheadsarewomen—butjustoneofour15divisionheads,andtwoofour21fullprofessors,arewomen.Theseproportionshavebarelychangedinthelastdecade,andthedis-paritieswereevengreaterfurtherback.Ofcoursestatisticsalonedon’tprovethatthereisaproblem.TheintroductorycommentsbyProfessorLotteBailyn,thenchairoftheMITfaculty,toalandmark1999reportputitnicely:Our first instinct is to deny that a problem exists (if it existed, it would surely have been solved by now), or to blame it on the pipeline or the circumstances and choices of individual women. Shecontinued, None of these, however, explains the inequities surfaced by the Committee.
Whydoesitmatter?ProfessorRachelWebsteroftheUniversityofMelbourne’sphysicsdepartmentsaidthis,atarecentGEworkshop:It is abundantly clear that women and men are equally talented in their capacity for scientific research. However they may not operate in the same way within our social context. Our challenge is to create
the environment that realises all that talent, rather than just a fraction of it.
Itcameasnosurprisetometolearnthat,tothewomenonourGEcommittee,thesinglemostimportantGEissueischild-care.Atlastourinstituteisdoingsome-thingaboutthat,althoughitwillclearlybesometimebeforeitisadequatelyaddressed.Iwassurprisedtolearnhowmuchisgoingonelsewhere topromoteGE,andembar-rassedtorealizethatIhavebeendraggingmyfeetforaquarterofacentury.
ManyUSuniversitiesnowhaveanofficeorcentreofWork/Life,tohelpimplementfamily-friendlypoliciesonparenting,child-andelder-care.Goodpoliciesontenureclockextension,modifieddutiesafterthearrivalofnewchildren,flexibleworkingpatterns,andemergencyassistance,arewidespread.Forexample,UCBerkeleygivesonesemesteroffteachingforallnewparents,andtwosemestersoffforthebirthparent.Thisgivesasemesterfortransitioningandresearchcatch-up,andasemesterforthephysicalmaternityleave.TheUniversityofMichigangivesbothtobothparents.TherearenowmanypoliciesonthesemattersinplacearoundtheUS,institutionalcontributionstoGE.
Inmyowninstitute,wenowtrytocon-finemeetingtimesto9:15–4:30;wehaveagenerouschild-caresubsidyforselectedindividuals;there’saprivateroomwheremotherscanbreast-feed,orexpressandstoremilk;afamilyroomforemergencychild-careisplanned;andwearetryingtomakeiteasiertoworkfromhome.Thecapacityexiststohiretemporarytechnicalassistanceforcontinuityofeffortduringmaternityleave.Wehavea“WomeninScience”lectureseries.Wesendwomenonleadership-trainingcoursesandwefundafamily-friendlyviewingroomataseriesofannualnationalconferences,
ChangesofthekindIhavejust
mentionedaretheeasiesttobringabout.Theideasmightcomefromthosebelow,buttheirimplementationisfromthetop.Andnotethattheytendtobeaimedatwomeninthefirst10–15yearsoftheircareers,andspecificallyaddresstheissueofchildren.Differentpoliciesandpracticesareneededtoaffectthemiddleandlateryearsofawoman’scareer,andtoincludewomenwhodonothavechildren.
Howcanwebringaboutchangestoinstitutionalandbroaderprofessionalcul-tures,toensurethatallwomenrealizetheirpotential?Therearenoquickfixeshere.Committeesaddressinghiring,prizes,con-ferenceprograms,andotherdecisionsmustconsciouslylookforwomenasfellowcom-mitteemembers,candidatesorspeakers,sincequalifiedfemalescanbeoverlookedonallfronts.Effectiveanduniformpoliciesformentoringalljuniorcolleaguesareneeded.Womenmayrequiremoresupportastheyapproachkeycareertransitionpoints,byencouragementtoapplyforpositionsorfellowships,andbyhaving“champions”inthesystemtopromotetheirinterests,e.g.throughanetworkofseniorcolleagues.Theremustbeequitabledistributionofrewards,resourcesandresponsibilities—salaries,fellowships,awards,space,teachingandcommitteeassignments—andeffectivemechanismstonoticeandaddressinequitieswhentheyarise.
Ibelievegenderequityis achievable,butgettingtherewillrequireimagina-tiveandsustainedeffortonthepartofallofus.
Terence’s Stuff: GenderEquityTerry Speed believes that gender equity is not just desirable but possible and beneficial—but it is not an easy journey
14 .IMsBulletin Volume40. Issue 5
At a glance:
forthcoming IMS Annual Meeting and JSM dates
2011IMS Annual Meeting @
JSM: MiamiBeach,FL,July30–August4,2011
2012IMS Annual Meeting
@ World Congress:
İstanbul,Turkey,July9–14,2012
JSM: SanDiego,CA,July28–August2,2012
2013IMS Annual Meeting
@ JSM: Montréal,Canada,August3–8,2013
2014IMS Annual Meeting:
Sydney,Australia,July7–11,2014
JSM: Boston,MA,August2–7,2014
2015IMS Annual Meeting
@ JSM: Seattle,WA,August8–13,2015
IMS meetings around the worldIMSco-sponsoredmeeting
NEWInternational Conference Ars Conjectandi 1713–2013October 15–16, 2013Basel, Switzerlandwhttp://www.statoo.ch/bernoulli13/
2013marksthe300thanniversaryofthepublicationofJacobBernoulli’sbook,Ars Conjectandi,in1713.Ameet-inghasbeenorganizedtocelebratethis:the“InternationalConferenceArs Conjectandi 1713–2013”willbeheldOctober15–16,2013,inBasel,Switzerland.
IMSRepresentativesontheprogramcommitteeareHansKünschandLutzDümbgen.
IMSsponsoredmeeting
IMS Annual Meeting @ 2011 Joint Statistical MeetingsJuly 30 – August 4, 2011, Miami Beach, FLw http://amstat.org/meetings/jsm/2011/
IMSsponsoredmeeting
IMS Annual Meeting @ 2013 Joint Statistical MeetingsAugust 3–8, 2013, Montréal, Quebec, Canadaw TBC
IMSsponsoredmeeting
2014 Joint Statistical MeetingsAugust 2–7, 2014, Boston, Mass., USAw TBC
IMSsponsoredmeeting
2012 Joint Statistical MeetingsJuly 28 – August 2, 2012, San Diego, CAw http://amstat.org/meetings/jsm/2012/
JointStatisticalMeetings,2011–2015
IMSAnnualMeetings,2012&2014
IMSsponsored
meeting
IMS Annual Meeting @ 2015 Joint Statistical MeetingsAugust 8–13, 2015Seattle, Washington, USAw TBC
IMSsponsoredmeeting
2012 World Congress/IMS Annual MeetingJuly 9–14, 2012Grand Cevahir Hotel & Convention Center, Istanbul, Turkeyw http://www.worldcong2012.org/
TheeighthWorldCongressinProbabilityandStatisticswillbeheldinIstanbulfromJuly9to14,2012.ItisjointlyorganizedbytheBernoulliSocietyandtheInstituteofMathematicalStatistics.Scheduledeveryfouryears,thismeetingisamajorworldwideeventforstatisticsandprobability,coveringallitsbranches,includingtheoretical,method-ological,appliedandcomputationalstatisticsandprobability,andstochasticprocesses.Itfeaturesthelatestscientificdevelopmentsinthesefields.
Contacts:ElvanCeyhanandMineÇağlar,Co-chairsoftheLocalOrganizingCommittee;ArnoldoFrigessi,ChairoftheProgramCommittee.
IMSsponsored
meeting
2014 IMS Annual MeetingJuly 7–11, 2014Sydney, Australiaw TBC
Thelocationforthe2014IMSAnnualMeetinghasbeenselectedasSydney,Australia.Detailswillfollow,butyoucanmarkyourcalendarsnow!
Istanbul’s Bosphorus Bridge connects Europe (on the left) and Asia (right)
August . 2011 IMsBulletin . 15
The Second IMS Asia Pacific Rim MeetingJuly 1–4, 2012Tsukuba, Japanw http://www.ims-aprm2012.org/
ProgramChairs:ByeongU.Parke [email protected]),RunzeLie [email protected]
Meetingpostponedfrom2011
SincethemassiveearthquakestruckJapanonMarch11,thelocalorganizingcommit-teeandthescientificprogramcommitteedecidedtopostponethemeetinguntilnextyear.WehaverescheduledittoJuly1–4,2012,andmovedittotsukuba,thesciencecityandacademiccenterofJapan,whichisabout60kmfromTokyo.
Weherebycordiallyinviteyoualltoattendthemeetingnextyear,whenwearecertainthatyouwillwitnessastrongrecoveryofJapanfromoneofthemostseverenaturaldisastersinrecenthistory.
Akimichi Takemura, LOC ChairByeong Park & Runze Li, SC Co-Chairs
IMSsponsoredmeeting
2013 ENAR/IMS Spring MeetingMarch 10–13, 2013Orlando, Florida, USAw http://www.enar.org/meetings.cfm
IMSsponsoredmeeting
2014 ENAR/IMS Spring MeetingMarch 16–19, 2014Baltimore, Maryland, USAw http://www.enar.org/meetings.cfm
IMSsponsoredmeeting
2012 ENAR/IMS Spring MeetingApril 1–4, 2012Washington DC, USAw http://www.enar.org/meetings.cfm
IMSco-sponsoredmeeting
International Symposium in Statistics (ISS) on Longitudinal Data Analysis Subject to Outliers, Measurement Errors, and/or Missing ValuesJuly 16–18, 2012Memorial University, St. John’s, Canadaw www.iss-2012-stjohns.ca
IMSRep:BrajendraSutradhar
IMSco-sponsoredmeeting
36th Conference on Stochastic Processes and their ApplicationsJuly 29 – August 2, 2013University of Colorado, Boulder, USAw TBC
IMSco-sponsoredmeeting
International Statistics Conference 2011December 28–30, 2011Colombo, Sri Lankaw TBC
OrganizedbytheAppliedStatisticsAssociationofSriLanka(ASASL).Themeetinglocationisatthewater’sedgeinthecapitalcityofSriLanka.IMSRep:PeterHall.Thewebsiteisunderconstruction.
IMSco-sponsoredmeeting
Colloquium in honor of Hans Rudolf Künsch on the occasion of his 60th birthdayOctober 3–4, 2011, ETH Zurich, SwitzerlandIMSReps:PeterBühlmann,MarloesMaathuis,SaravandeGeerw https://stat.ethz.ch/events/Colloquium_Kuensch
KeynotespeakersareJimBerger(DukeUniversity),StuartGeman(BrownUniversity),PeterGreen(UniversityofBristol).Invitedspeakersare:RainerDahlhaus(UniversityofHeidelberg),ArnoldoFrigessi(UniversityofOslo),ReinhardFurrer(UniversityofZurich),HavardRue(NorwegianUniv.S&TTrondheim),RetoKnutti(ETHZurich),ChristianP.Robert(UniversitéParis-Dauphine).
Hans R. Künsch
ENAR,2012–2014
16 .IMsBulletin Volume40. Issue 5
Other meetings around the worldConcentration inequalities and their applicationsJanuary 23–27, 2012Centre International de Rencontres Mathématiques (CIRM), Marseille, Francew http://www.cirm.univ-mrs.fr/
Theworkshopwillfocusonrecentadvancesandapplicationsofmeasureconcentrationandwillgatherparticipantsfromvari-ousareaswherethisideahasbeenusefulandfruitfultofavourexchangesandcollaborativeresearch.Thetopicsofinterestare,amongothers,asymptoticgeometricanalysis,probabilitytheory,statisticalmechanics,randommatrices,dynamicalsystems,(quan-tum)informationtheory,randomgraphs,randomizedalgorithms,statistics,compressedsensing,learningtheory.
Theworkshopwillbeorganizedaroundfourmini-coursesbyS.Chatterjee,CourantInstitute,USA;A.Guillin,UniversityofClermont-Ferrand,France;P.Massart,UniversityofParisXIatOrsay,France;andA.Panconesi(LaSapienzaUniversity,Roma)&D.Dubhashi(UniversityofTrier,Germany).Therewillalsobeinvitedtalksondifferenttopicscoveringthemaininterests.
TheCIRMcoversallthelivingexpensesbutnottravelexpenses.Lookingforwardtoyourparticipationtotheworkshop.Theorganizers:J.-R.Chazottes,EcolePolytechnique,Paris,FranceM.Ledoux,UniversityofToulouse,FranceF.Redig,RadboudUniversityofNijmegen,TheNetherlands
Analytical Methods in Statistics (AMISTAT)October 28–30, 2011Prague, Czech Republicw http://amistat2011.karlin.mff.cuni.cz
TheWorkshopAnalyticalMethodsinStatistics(AMISTAT)isheldinhonorofProfessorAbramKaganinPraguefromOctober28toOctober30,2011.ThemainspeakersareAbramKagan,IldarIbragimov,LevB.Klebanov,GerardLetac,LutzMattner,YakovNikitin,StephenPortnoy,YosefRinott,GeorgyShevlyakov.Discussant:DmitriiChibisov.Contact:[email protected]
orJanaJureckova:[email protected]
SecondAnnouncement
Third Latin American Meeting on Bayesian Statistics(3er Congreso Bayesiano de América Latina—COBAL)and XXXVIII National Meeting of Statistics(XXXVIII Jornadas Nacionales de Estadística—JNE)October 23–27, 2011Pucón, Chilew http://cobal2011.usach.cl
The3rdLatinAmericanMeetingonBayesianStatisticsandtheXXXVIIINational(Chilean)MeetingofStatistics,organizedbytheUniversidaddeSantiagodeChileincollaborationwithseveralnationaluniversities,willtakeplaceonOctober23–27,2011,attheexoticandbeautifulsoutherncityofPucón,Chile(http://www.
visitpucon.com).Themaingoalsofthisjointeventaretofacilitatetheexchange
ofrecentresearchdevelopmentsinstatisticalmethodologyandtheoreticaldevelopments(BayesianforCOBAL,miscellaneousforJNE),toestablishnewcollaborationsandpartnershipsamongattendants,andfinally,toprovideopportunitiesforstudentsandresearcherstoshareideasanddiscussstatisticalmethods.
ConfirmedInvitedSpeakersareTrevorHastie(StanfordUniversity,USA);PeterMüller(UniversityofTexas,Austin,USA);FabrizioRuggeri(CNR-IMATI,Italia);WilfredoPalma(PontificiaUniversidadCatólicadeChile);FernandoQuintana(PontificiaUniversidadCatólicadeChile);Jean-MichelMarin(UniversitéMontpellierII,France);AlexandraSchmidt(UniversidadeFederaldoRiodeJaneiro,Brasil);RamsésMena(UniversidadNacionalAutónomadeMéxico);andSujitGhosh(NorthCarolinaStateUniversity,USA).
TherewillbethematicsessionsrelatedtoIdentifiability,BiostatisticsandBayesianFoundations,aswellasanAMSBI(AppliedStochasticModelinginBusinessandIndustry)discussionpapersession.
Aconfirmedmini-courserelatestospecifictopicson“ItemResponseDataModeling”.Tentatively,thesecondoneisentitled“MeasurementErrorModeling”
Thethematicsessionswillincludeayoungstatisticianssessionfordoctoralstudentsinthefinalstageoftheirthesiswork,oryoungresearchers,whowanttoshareandshowtheirmorerecentdevelopments.Therewillalsobeoralcommunicationsandpostersessionsforcontributedpapers.
Registrationisavailableonthewebsite.Forforeignattendants,itisimportanttoberegisteredbeforethedeadline.Theymaymaketheirpaymentsthefirstdayofthemeeting.Seewebsitefordeadlines
Meetingemaile [email protected]
WearelookingforwardtoseeingyouinPucón!OrganizingCommittee
August . 2011 IMsBulletin . 17
Statistical Analyses for Next Generation Sequencing September 26–27, 2011Birmingham, Alabamaw http://www.soph.uab.edu/ssg/courses/nhgri_r13/ngsstat
TheUniversityofAlabamaatBirmingham’sSectiononStatisticalGeneticsispleasedtoannounceourNHGRI-fundedconferenceonStatisticalAnalysesforNextGenerationSequencing.
Next-generationsequencingtechnologyisimpactingalmostallaspectsofbiomedicalresearch.Thistechnologygeneratesanunprecedentedwealthofdatathatdemandsnovelanalysisstrate-gies.WhileITinfrastructureandbioinformaticsdevelopmentsareobviouslyrequiredtoenablesoundinformationextraction,sophisticatedstatisticalmethodologiesandalgorithmsarealsoessentialforinterpretingthedata.Inthisregard,weareorganizingaNHGRIfundedtwo-dayconference,callingstatisticians,geneticepidemiologists,bioinformaticians,andgenomebiologists,todis-cussthestatisticalchallengesandopportunitiesinnext-generationsequencingdataanalysis.Webelievethatthisconferencewillprovideavenueforexchangingofcutting-edgeinformationandideas,andfosteringcollaborationsamongmethodologists,analysts,andbiomedicalinvestigators.Topicsinclude:1. Newadvancesintheapplicationofnext-generationsequencing
technology2. RNA-seqandmiRNA-seqdataanalyses3. ChIP-seq,methyl-seqdataanalyses,andintegrationwithother
highdimensiondatatypes.4. Next-generationsequencingfordenovosequencingandrese-
quencing5. Incorporatingnext-generationsequencingdataintothegenetics
studyofcomplexdisease6. RareVariantAssociationAnalysesandnextgenerationsequenc-
ing7. SNPandstructurevariationanalysisusingnext-generation
sequencing8. Haplotypeanalysiswithsequencingdata
PosterSessionavailable!DeadlineforAbstractsubmissionforpostersandshorttalksisAugust14,2011.ShorttalkacceptancewillbenotifiedbyFriday,Aug21,2011.
InvitedSpeakers:HongkaiJi,SuzanneLeal,Xiaole(Shirley)Liu,AliMortazavi,RichardMyers,BenjaminRaphael,StephanieSantorico,HongyuZhao,ThomasWu.
Attendanceislimited.Pleaseregisterearly.Women,membersofunderrepresentedminoritygroupsandindividualswithdisabilitiesarestronglyencouragedtoapply.
NIMBioS Investigative Workshop: Mathematical Modeling of Intracellular MovementsOctober 24–26, 2011Knoxville, TNw http://nimbios.org/workshops/WS_intracellular_mv.html
TheNationalInstituteforMathematicalandBiologicalSynthesis(NIMBioS)isnowacceptingapplicationsforitsInvestigativeWorkshop,“MathematicalModelingofIntracellularMovements”tobeheldOctober24–26,2011,atNIMBioS.
Recentadvancesinlivecellmicroscopyhaveresultedinafloodoftime-lapseobservationsthatrevealahighdegreeofmotilityinsidecells.Quantitativeanalysisofthesemovementsisnecessarytogainafullunderstandingofintracellulardynamicsandtheirregulation.Thisanalysisisoftenhamperedbythesheercomplexityofthemovements,thegreatnumberofobjectstobetracked,andthediffractionlimitofopticalmicroscopes.Atthesametime,mathematicalandstatisticalmodelshavemadesignificantprogressinproducingfastalgorithmsthatreliablytrackmultipleobjectsinspace.Insomecases,thesemodelsweresuccessfullyappliedtocellbiologicaldatasets,butthefullpotentialofarigorousmathemati-calapproachthatcanbeemployedacrossawiderangeofbiologicalprocesseshasnotbeenrealized.ThisNIMBioSWorkshopwillbringtogetherexpertsfromcellbiologyaswellasmathematics,statistics,computationalscienceandphysicstodiscusscurrentapproachesandpossiblealternatives.
Formoreinformationabouttheworkshopandalinktotheonlineapplicationform,gotothewebsiteabove.
Ifneeded,applicantsmayrequesttravelandlodgingsupport.Participationislimited,andthoseselectedtoattendwillbenotifiedwithintwoweeksoftheapplicationdeadline.Applicationdeadline:July 24, 2011
The 2012 Stochastic Networks ConferenceJune 18–22, 2012MIT, Cambridge, MA, USAw http://stoch-nets-2012.lids.mit.edu/
Thiseventwillbringtogetherresearcherswhoshareaninterestinstochasticnetworkmodels,tosurveyrecentdevelopments,andidentifyfutureresearchdirections.Themeetingwillbestructuredinaworkshopformat,withapproximatelytwentyhour-longinvitedtalks,andamplefreetime,soastomaximizeinteractionsbetweenspeakersandparticipants,andtofacilitateafruitfulexchangeofideas.Inaddition,therewillbeapostersessionforcontributedpapers.
NEWNEW
NEW
18 .IMsBulletin Volume40. Issue 5
::: Advertise current job opportunities for only $250 for 60 days ::: See http://jobs.imstat.org for details :::
Canada: Montreal, QuébecMcGill University Faculty of MedicineChairoftheDepartmentofEpidemiology,BiostatisticsandOccupationalHealthhttp://jobs.imstat.org/c/job.cfm?site_id=1847&jb=8294487
Canada: Waterloo, OntarioUniversity of WaterlooPostdoctoralFellowshiphttp://jobs.imstat.org/c/job.cfm?site_id=1847&jb=8172622
Canada: Waterloo, OntarioUniversity of Waterloo, Department of Statistics and Actuarial ScienceChairofStatisticsandActuarialSciencehttp://jobs.imstat.org/c/job.cfm?site_id=1847&jb=8079833
Canada: Waterloo, OntarioUniversity of WaterlooChairofStatisticsandActuarialSciencehttp://jobs.imstat.org/c/job.cfm?site_id=1847&jb=8069684
Mexico: GuanajuatoCIMATDepartment of Probability and StatisticsTwopositionsSeeadvertisementonnextpage
New Zealand: Dunedin and Palmerston North, OtagoAgResearchStatistician/SeniorStatisticianhttp://jobs.imstat.org/c/job.cfm?site_id=1847&jb=8062587
United Kingdom: LondonUniversity College LondonChairsofStatisticsx2http://jobs.imstat.org/c/job.cfm?site_id=1847&jb=8032140
TheDepartmentofMathematicsatCornellUniversityinvitesapplicationsfortwotenure-trackAssistantProfessorpositions,orhigherrank,pendingadministrativeapproval,startingJuly1,2012.ThesearchesareopentoallareasofMathematicswithanemphasisontheareasofprobability,numbertheory,andPDE.TheDepartmentactivelyencouragesapplicationsfromwomenandminoritycandidates.
Applicantsmustapplyelectronicallyathttp://www.mathjobs.
org.
Forinformationaboutourpositionsandapplicationinstruc-tions,see:http://www.math.cornell.edu/Positions/positions.html
Applicantswillbeautomaticallyconsideredforalleligibleposi-tions.
DeadlineNovember1,2011.Earlyapplicationswillberegardedfavorably.
CornellUniversityisanAffirmativeAction/EqualOpportunityEmployerandEducator.
Employment Opportunities around the world
United States: Ithaca, New YorkCornell UniversityDepartment of MathematicsTwotenure-trackAssistantProfessorpositions,orhigherrankSeeadvertisementbelow
Wehave
stre
tched
theminimum
timeforwhicha
singlejobpostingis
displayedonline.
Single60-dayjobposting $250.00
Single90-dayjobposting $365.00
Single120-day jobposting $480.00
Single150-day jobposting $595.00
Placeyourjobadathttp://jobs.imstat.org/
August . 2011 IMsBulletin . 19
::: Search our online database of the latest jobs around the world for free at http://jobs.imstat.org :::
Mexico: GuanajuatoPositions in Statistics at Centro de Investigación en Matemáticas (CIMAT), Guanajuato, Mexico CentrodeInvestigaciónenMatemáticas(CIMAT),locatedinGuanajuato,Mexico(www.cimat.mx),isafederallyfundedCONACYT(www.conacyt.gob.mx)researchcenterdevotedtoresearch,teachingandapplications.Itcomprisesdepartmentsinpureandappliedmathematics,probabilityandstatistics,andcomputerscience.Undergraduateprogramsinmathematics,andgraduate(master’sanddoctoral)programsinallrepresentedbranchesarefostered.SomeCIMATsignaturesaredynamicresearchgroupsinboththeoreticalandappliedfields,recruitmentofsomeofMexico’sfinestmathematicsstudents,atime-honoredtraditioninhostingnumerousaca-demicnationalandinternationalmeetings,consultingwithscientistsandindustryandgovernment,outreachprogramsforyoungsters,andtheadvancementofmathematicsandscienceinsociety.
TheDepartmentofProbabilityandStatistics(www.cimat.mx/index.php?m=3)invitesapplicationsfortwopositions,beginningasearlyasJanuary,2012.Allareasofstatisticsaresought,withpreferencegiventocurrent,challengingandimportanttopicswithpoten-tialforgroupdevelopmentandmultidisciplinarywork.
ApplicantsarerequiredtoholdaPh.D.degreeinstatisticsbythetimeofemployment.Dutiesconsistofresearchandgraduateandundergraduatelevelqualityteachinginamathematicallyinclinedcurriculum,forwhichtheSpanishlanguageisanasset,yetnotanabsoluterequirement.TheDepartmenthostsaMaster’sandDoctoralprograminProbabilityandStatistics.
Atotaloftwopositionsareavailable,inanycombinationofthefollowingthreetypes:
A)Postdoctoral, or young researcher position,foramaximumofthreeyears.Specificrequirements:i. Demonstratepotentialforimportantacademicresearchinstatistics.ii. WillingnesstoapplytotheSistemaNacionaldeInvestigadores(SNI)(www.conacyt.mx/SNI/Index_SNI.html,inSpanish),
afederal-funded,peer-reviewedresearchgrantagency.
B)Experienced researcher.Requirements:i. Anoteworthycareerinresearch,havingproductivityconsistentwithtimesincegraduation.ii. AcurriculumvitaethatisagreeableforapplicationtotheSNI(seeabove),atalevelconsistentwithtimesincegraduation.iii. Potentialforleadership.iv. Teachingexperienceinintroductoryandadvancedsubject-mattercoursesinundergraduateandgraduateprogramssupported
bytheDepartmentofProbabilityandStatisticsatCIMAT.
C)Sabbatical stay, or visitor position. Inadditiontotherequirementsdescribedaboveforoption(B),potentialforthestrengtheningofacademicprogramsandresearchattheDepartmentofProbabilityandStatistics.
InterestedapplicantsshouldsendaletterofapplicationtoDr.DanielHernándezHernández([email protected]),beforeNovember
30,2011,withthefollowingmaterialsattached:1)CurrentCurriculumvitae.2)Briefdescriptionofresearchtopics.3)WorkplanspanningthefirstyearatCIMAT.4)Twolettersofreference.5)Dateoflikelyincorporationandtypeofpositionsought.
20 .IMsBulletin Volume40. Issue 5
International Calendar of Statistical EventsIMS meetings are highlighted in maroonwith the logo, and new or updated entries have the NEW or UPDATED symbol. t means telephone, f fax, e email and w website. Please submit your meeting details and any corrections to Elyse Gustafson at [email protected]
July–August 2011
July30–August4:MiamiBeach,Florida.IMsAnnualMeetingatJsM2011.w http://amstat.org/meetings/jsm/2011/
July31&August3:atJSMMiami.2011nIss/AsAWritingWorkshopforJuniorResearchersw http://www.amstat.org/
meetings/wwjr/
August 2011
August1–4:Boulder,Colorado,USA.UncertaintyQuantificationinscientificComputing.w http://www.nist.gov/itl/math/ifip-
woco-10.cfm
August1–5:SandbjergEstate,Sønderborg,Denmark.ConferenceinHonourofsørenAsmussen:newFrontiersinAppliedProb-abilitywwww.thiele.au.dk/asmussen
August11–13:UniversityofConnecticut,Storrs,USA.46thActuarialResearchConference.w http://www.math.uconn.
edu/~valdez/46arc/46arc-storrs.php
August15–19:Prague,CzechRepublic.nonparametricsandGeometrywhttp://nonparam11.karlin.mff.cuni.cz
August16–18:UniversityofWarwick,Coventry,UK.useR!2011w http://www.R-project.org/useR-2011
August17–19:Belfast,NorthernIreland.InternationalAssociationforofficialstatisticsconference“The Demography of Ageing and Official Statistics” w http://www.nisra.gov.uk/IAOS2011.html
August17–19:Copenhagen,Denmark.DynamicstatisticalModels[ISISatelliteMeeting]w http://statistics.ku.dk/isi-satellite/
August19–21:Dublin,Ireland.YoungstatisticiansMeeting(YsI2011)ISI Satellite Meeting w http://www.scss.tcd.ie/conferences/
YSI2011
August21–26:Dublin,Ireland.IsIDublin:58thWorldstatisticsCongressw www.isi2011.ie
August29–September1:WashingtonDC,USA.7thInternationalConferenceonMultipleComparisonProceduresw http://www.mcp-conference.org
September 2011
September5–9:Lisbon,Portugal.17theuropeanYoungstatisticiansMeetingwhttp://www.fct.unl.pt/17eysm
September7–8:StatisticalCenterofStatisticsKorea,Daejeon,SouthKorea.ThirdInternationalWorkshoponInternetsurveyMethodswwww.kostat.go.kr
September12–16:MathematicalBiosciencesInstitute,Columbus,Ohio.WorkshoponnewQuestionsinProbabilityTheoryArisinginBiologicalsystemsw http://mbi.osu.edu/2011/ws1description.html
September12–December16:InstituteforPureandAppliedMathematics,LosAngeles,USA.MathematicalandComputationalApproachesinHigh-ThroughputGenomicsw www.ipam.ucla.edu/programs/gen2011/
September13–16:Jaca,Spain.statistics,ProbabilityandoperationsResearch(sPo2011)w http://metodosestadisticos.
unizar.es/~jaca2011
September24:Cambridge,MA.2011newenglandsymposiumonstatisticsinsportswhttp://www.amstat.org/chapters/boston/
nessis11.html
September25–28:Ribno,Bled,Slovenia.Appliedstatistics2011whttp://conferences.nib.si/AS2011
NEW September26–27:Birmingham,Alabama.statisticalAnalysesfornextGenerationsequencingw http://www.soph.uab.
edu/ssg/courses/nhgri_r13/ngsstat
October 2011
October3–4:ETHZurich,Switzerland.ColloquiuminhonorofHansRudolfKünschontheoccasionofhis60thbirth-dayw https://stat.ethz.ch/events/Colloquium_Kuensch
October12–14:WashingtonDC/SilverSpringMD.ConferenceonRiskAssessmentandevaluationofPredictionswhttp://brac.
umd.edu/~Risk2011/Main.htm
October18–20:HarvardMedicalSchool,Cambridge,Mass.2011non-clinicalBiostatisticsConferencew http://www.hsph.harvard.
edu/ncb2011/
August . 2011 IMsBulletin . 21
NEW October23–27:Pucón,Chile.ThirdLatinAmericanMeetingonBayesianstatistics(CoBAL)andXXXVIIIChileannationalMeetingofstatistics(Jne)w http://cobal2011.usach.cl
NEW October24–26:Knoxville,TN,USA.nIMBiosInvestigativeWorkshop:MathematicalModelingofIntracellularMovementsw http://nimbios.org/workshops/WS_intracellular_
mv.html
NEW October28–30:Prague,CzechRepublic.AnalyticalMethodsinstatistics(AMIstAt)w http://amistat2011.karlin.mff.cuni.cz
November 2011
November7–9:Łódź,Poland.MultivariatestatisticalAnalysisConferencewhttp://www.msa.uni.lodz.pl
November25–27:Lahore,Pakistan.3rdInternationalConferenceonstatisticalscienceswhttp://www.icss3.co.nr/
December 2011
December18–21:Amman,Jordan.11thIslamicCountriesConferenceonstatisticalsciences(ICCs-11)whttp://www.iccs11.
isoss.net/
December28–30:Colombo,SriLanka.InternationalstatisticsConference2011.w TBC
December28–31:HongKong,China.InternationalConferenceonAdvancesinProbabilityandstatisticsTheoryandApplica-tions:Acelebrationofn.Balakrishnan’[email protected]://faculty.smu.edu/ngh/
icaps2011.html
January 2012NEW January1–6:Hyderabad,India.22ndAnnualConferenceof
TheInternationalenvironmetricssocietyw www.ties2012.com/
NEW January23–27:CentreInternationaldeRencontresMathématiques(CIRM),Marseille,France.Concentrationinequalitiesandtheirapplicationsw http://www.cirm.univ-mrs.fr/
April 2012
April1–4:WashingtonDC,USA.2012enAR/IMsspringMeetings.w http://www.enar.org/meetings.cfm
April18–20:Poznań,Poland.InternationalCongressofPolishstatisticstocelebratethe100thanniversaryofthePolishStatisticalAssociationw http://www.stat.gov.pl/pts/
March 2012 NEW March14–16:HongKong.IAenGInternationalConference
onDataMiningandApplications2012w www.iaeng.org/
IMECS2012/ICDMA2012.html
NEW March30–31:WashingtonDC.Informationandeconometricsofnetworksw www.american.edu/cas/economics/
info-metrics/workshop/workshop-2012-spring.cfm
June 2012
June3–6:Guelph,Ontario,Canada.ssCAnnualMeetingw TBC
NEW June18–22:MIT,Cambridge,MA,USA.The2012stochasticnetworksConferencew http://stoch-nets-2012.lids.mit.
edu/
June23–26:Boston,MA,USA.ICsA2012Appliedstatisticssymposium.w TBC
June25–29:Kyoto,Japan.2012IsBAWorldMeetingw http://www2.e.u-tokyo.ac.jp/~isba2012/
July 2012
July1–4:Tsukuba,Japan.IMsAsiaPacificRimMeetings.w
http://www.ims-aprm2012.org/(meeting postponed from July 2011
due to the earthquake)
July3–6:UniversityofOslo,Norway.ThirdbiennialInternationalstatisticalecologyConferencew http://www.cees.uio.no/news/2010/isec2012.html
Continuesonpage 22
22 .IMsBulletin Volume40. Issue 5
July9–14:Istanbul,Turkey.IMsAnnualMeeting2012inconjunctionwith8thWorldCongressinProbabilityandstatistics.w http://www.worldcong2012.org/
July16–18:MemorialUniversity,St.John’s,Canada.Inter-nationalsymposiuminstatistics(Iss)onLongitudinalDataAnalysissubjecttooutliers,Measurementerrors,and/orMissingValuesw www.iss-2012-stjohns.ca
July28–August2:SanDiego,California.JsM2012.w http://amstat.org/meetings/jsm/2012/index.cfm
March 2013
March10–13:Orlando,Florida.2013enAR/IMsspringMeeting.w http://www.enar.org/meetings.cfm
July 2013
July29–August2:UniversityofColorado,Boulder,USA.36thConferenceonstochasticProcessesandtheirApplicationsw
TBC
August 2013
August3–8:Montréal,Canada.IMsAnnualMeetingatJsM2013.w http://amstat.org/meetings/jsm/
August24–31:HongKong.InternationalstatisticalInstitute:59thIsIWorldstatisticsCongressw www.isi2013.hk
October 2013NEW October15–16: Basel,Switzerland.International
ConferenceArs Conjectandi1713–2013w http://www.statoo.ch/
bernoulli13/
March 2014
March16–19:Baltimore,Maryland.2014enAR/IMsspringMeeting. w http://www.enar.org/meetings.cfm
July 2014
July7–11:Sydney,Australia.2014IMsAnnualMeeting.w TBC
August 2014
August2–7:Boston,MA.JsM2014andAsA’s175thAnniversary.w TBC
August 2015
August8–13:Seattle,WA.JsM2015.w TBC
Are we missing something? If you know of
any statistics or probability meetings which
aren’t listed here, please let us know. Email
the details to Elyse Gustafson at erg@imstat.
org. We’ll list them here in the
Bulletin, and online too, at
www.imstat.org/meetings
International Calendar continued
July 2012 continued
August . 2011 IMsBulletin . 23
Information for Advertisers General information: TheIMS BulletinandwebpagesaretheofficialnewsorgansoftheInstituteofMathematicalStatistics.TheIMS Bulletin,establishedin1972,ispublished8timesperyear.Printcirculationisaround2,000papercopies,anditisalsofreeonlineinPDFformatathttp://bulletin.imstat.org,postedonlineabouttwoweeksbeforemailout(averagedownloadsover8,000).SubscriptiontotheIMS Bulletincosts$90.Tosubscribe,call877-557-4674(UStoll-free)or+12162952340(international),oremailstaff@
imstat.org.TheIMSwebsite,http://imstat.org,establishedin1996,receivesover30,000visitspermonth.Publicaccessisfree.
Advertising job vacancies Asingle60-dayonlinejobpostingcostsjust$250.00.Wewillalsoincludethebasicinformationaboutyourjobad(positiontitle,location,companyname,jobfunctionandalinktothefullad)intheIMS Bulletinatnoextracharge.Seehttp://jobs.imstat.org
Advertising meetings, workshops and conferencesMeetingannouncementsintheBulletinandontheIMSwebsiteathttp://imstat.org/meetingsarefree.SendthemtoElyseGustafson;seehttp://www.imstat.org/program/prog_announce.htm
Rates and requirements for display advertisingDisplayadvertisingallowsforplacementofcamera-readyadsforjournals,books,software,etc.Acamera-readyadshouldbesentasagrayscalePDF/EPSwithallfontsembedded.EmailyouradverttoAudreyWeiss,[email protected]://bulletin.imstat.org/advertise
Dimensions:widthxheight Rate
1/3page 4.93"x4"(125.2x102mm) $215
1/2page 7.5"x4"(190x102mm) $270
2/3page 4.93"x8"(125.2x203mm) $325
Fullpage(toedge,including1/8"bleed)
8.75"x11.25"(222mmx285.8mm) $380
Fullpage(withinusualBulletin margins)
7.5"x9.42"(190mmx239.3mm) $380
Deadlines and Mail Dates for IMS Bulletin
Issue Deadline Onlineby Mailed
1: January/February December 1 December15 January12: March February 1 February15 March13: April/May March 15 April1 April154: June/July May 1 May15 June15: August July 1 July15 August16: September August 15 September1 September157: Oct/Nov September 15 October1 October158: December November 1 November15 December1
MembershipandsubscriptionInformation
Journals ThescientificjournalsoftheInstituteofMathematicalStatisticsareThe Annals of Statistics,The Annals of Probability, The Annals of AppliedStatistics,The Annals of Applied Probability,andStatistical Science.TheIMS BulletinisthenewsorganoftheInstitute.
Individual and Organizational Memberships EachindividualmemberreceivestheIMS Bulletin (printand/orelectronic) andmayelecttoreceiveoneormoreofthefivescientificjournals.Memberspayannualduesof$103.Anadditional$53isaddedtotheduesofmembersforeachscientificjournalselected.Reducedmembershipduesareavailabletofull-timestudents,newgraduates,permanentresidentsofcountriesdesignatedbytheIMSCouncil,andretiredmembers.organizationalmembershipsareavailabletodepartments,corporations,governmentagenciesandothersimilarresearchinstitutionsat$155peryear.
Individual and General Subscriptions Subscriptionsareavailableonacalendar-yearbasis.Individualsubscriptionsareforthepersonaluseofthesubscriberandmustbeinthenameof,paiddirectlyby,andmailedtoanindividual.Individualsubscriptionsfor2011areavailabletoThe Annals of Applied Probability($166),The Annals of AppliedStatistics ($166),The Annals of Probability ($166),The Annals of Statistics($166),Statistical Science ($146),andIMS Bulletin ($113).Generalsubscriptionsareforlibraries,institutions,andanymultiple-readershipuse.Generalsubscriptionsfor2011areavailabletoThe Annals of Applied Probability($360),The Annals of AppliedStatistics ($320),The Annals of Probability ($360),The Annals of Statistics($390),Statistical Science ($210),andIMS Bulletin ($95).AirmailratesfordeliveryoutsideNorthAmericaare$109pertitle.
IMS Bulletin TheIMS BulletinpublishesarticlesandnewsofinteresttoIMSmembersandtostatisticiansandprobabilistsingeneral,aswellasdetailsofIMSmeetingsandaninternationalcalendarofstatisticalevents.ViewsandopinionsineditorialsandarticlesarenottobeunderstoodasofficialexpressionsoftheInstitute’spolicyunlesssostated;publicationdoesnotnecessarilyimplyendorsementinanywayoftheopinionsexpressedtherein,andtheIMS Bulletin anditspublisherdonotacceptanyresponsibilityforthem.TheIMS Bulletin iscopyrightedandauthorsofindividualarticlesmaybeaskedtosignacopyrighttransfertotheIMSbeforepublication.
TheIMS Bulletin(ISSN1544-1881)ispublishedeighttimesperyearinJanuary/February,March,April/May,June/July,August,September,October/NovemberandDecember,bytheInstituteofMathematicalStatistics,3163SomersetDr,Cleveland,Ohio44122,USA.PeriodicalspostagepaidatCleveland,Ohio,andatadditionalmailingoffices.Postmaster:SendaddresschangestoInstituteofMathematicalStatistics,9650RockvillePike,SuiteL3503A,Bethesda,MD20814-3998.
Copyright©2011bytheInstituteofMathematicalStatistics.
PrintedbyTheSheridanPress,450FameAvenue,Hanover,PA17331,USA.
The purpose of the Institute is to foster the
development and dissemination
of the theory and applications of
statistics and probabilityIMS: Organized September 12, 1935
Volume40. Issue 5
ISSN (1050-5164)
THE ANNALSof
APPLIEDPROBABILITY
AN OFFICIAL JOURNAL OF THE
INSTITUTE OF MATHEMATICAL STATISTICS
Articles
Geodesics and flows in a Poissonian city . . . . . . . . . . . . . . . . . . . . . . . WILFRID S. KENDALL 801A rule of thumb for riffle shuffling
SAMI ASSAF, PERSI DIACONIS AND K. SOUNDARARAJAN 843General notions of indexability for queueing control and asset management
KEVIN D. GLAZEBROOK, DAVID J. HODGE AND CHRIS KIRKBRIDE 876Efficient simulation of nonlinear parabolic SPDEs with additive noise
ARNULF JENTZEN, PETER KLOEDEN AND GEORG WINKEL 908A positive recurrent reflecting Brownian motion with divergent fluid path
MAURY BRAMSON 951Connectivity and equilibrium in random games
CONSTANTINOS DASKALAKIS, ALEXANDROS G. DIMAKISAND ELCHANAN MOSSEL 987
Concentration inequalities for mean field particle modelsPIERRE DEL MORAL AND EMMANUEL RIO 1017
Hamilton cycles in random geometric graphsJÓZSEF BALOGH, BÉLA BOLLOBÁS, MICHAEL KRIVELEVICH,
TOBIAS MÜLLER AND MARK WALTERS 1053Bounds on the speed and on regeneration times for certain processes on regular trees
ANDREA COLLEVECCHIO AND TOM SCHMITZ 1073Recovering a time-homogeneous stock price process from perpetual option prices
ERIK EKSTRÖM AND DAVID HOBSON 1102Gaussianization and eigenvalue statistics for random quantum channels (III)
BENOÎT COLLINS AND ION NECHITA 1136Analysis of market weights under volatility-stabilized market models . . . . . . SOUMIK PAL 1180
Vol. 21, No. 3—June 2011
nextthe
issue is
September 2011
DEADLINESfor
submissions
Journalalerts
August 15, then September 15Pleaseseeinsidethebackcoverforsubscriptiondetailsandinformationforadvertisers,includingallourdeadlinesandrequirements
ForalertsandspecialinformationonalltheIMSjournals,signupattheIMSListssitehttp://lists.imstat.org
Meetingreports,newsofmembers,informationandannouncementsaboutconferences,andjobsaroundtheworld.
Sendinyourideas,articles,andletters…Welovetohearfromyou!
AAP June 2011
http://projecteuclid.org/aoap