|
|
|
- Alaina Farmer
- 10 years ago
- Views:
Transcription
1 NoFreeLunchTheoremsforSearch SFI-TR TheSantaFeInstitute 1399HydeParkRoad SantaFe,NM,87501 February23,1996 possiblecostfunctions.inparticular,ifalgorithmaoutperformsalgorithmbonsome formexactlythesame,accordingtoanyperformancemeasure,whenaveragedoverall Weshowthatallalgorithmsthatsearchforanextremumofacostfunctionper- Abstract whereboutperformsa.startingfromthisweanalyzeanumberoftheotherapriori characteristicsofthesearchproblem,likeitsgeometryanditsinformation-theoretic aspects.thisanalysisallowsustoderivemathematicalbenchmarksforassessinga costfunctions,thenlooselyspeakingtheremustexistexactlyasmanyotherfunctions time-varyingcostfunctions.weconcludewithsomediscussionofthejustiabilityof particularsearchalgorithm'sperformance.wealsoinvestigateminimaxaspectsof biologically-inspiredsearchmethods. functiontopredictfuturebehaviorofthesearchalgorithmonthatcostfunction,and thesearchproblem,thevalidityofusingcharacteristicsofapartialsearchoveracost 1Manyproblemscanbecastasoptimizationovera\cost"or\tness"function.Insucha problem,wearegivensuchafunction,f:x!y(fbeingthesetofallsuchmappings). Introduction Physicalexamplesofsuchaproblemincludefreeenergyminimization(Y=<)overspin congurations(x=f 1;+1gN),oroverbondangles(X=f<<<gN),etc.Examplesalsoaboundincombinatorialoptimization,rangingfromnumberpartitioningtograph weseekthex'swhichextremizef(thiswilloftenbeimplicitlyassumedinthispaper). Forthatfweseekthesetofx2Xwhichgiverisetoaparticulary2Y.Mostoften, coloringtoscheduling[4]. 1
2 tematicconstructionofagoodxvalue,x0,fromgoodsub-solutionsspecifyingpartofx0. Themostcelebratedmethodofthistypeisthebranchandboundalgorithm[9].Forthis systematicandexhaustiveapproachtoworkinreasonabletime,onemusthaveaneective Therearetwocommonapproachestotheseoptimizationproblems.Therstisasys- work[11]linkingthecostfunctiontothepropertiesaheuristicmusthaveinordertosearch heuristic,h(n),representingthequalityofsub-solutionsn.thereisextensivetheoretical eciently. [7],andgeneticalgorithms[5]. values.therearemanyalgorithmsofthistype,includinghill-climbing,simulatedannealing solutionsx2xandtheassociatedyvalues,and(triesto)iterativelyimprovesuponthosex Asecondapproachtooptimizationbeginswithapopulationofoneormorecomplete thesealgorithmsaredirectlyapplied,withlittleornomodication,toanycostfunctionina biasesinhowtheytrytoimprovethepopulation(i.e.,thebiasesinhowtheysearchx) must\match"thoseimplicitinthecostfunctiontheyareoptimizing.howeveralmostalways Intuitively,onewouldexpectthatforthisclassofalgorithmstoworkeectively,the wideclassofcostfunctions.theparticularsofthecostfunctionsathandarealmostalways broadclassofproblemsisrarelyjustied. thecostfunctionarecrucial,andblindfaithinanalgorithmtosearcheectivelyacrossa ignored.aswewilldemonstratethough,the\matching"intuitionistrue;theparticularsof onemightexpectthathill-climbingusuallyoutperformshill-descendingifone'sgoalisto ndamaximumofthecostfunction.onemightalsoexpectitwouldoutperformarandom AperformsbetterthanBonaverage,evenifBsometimesoutperformsA.Asanexample, Indeed,onemightexpectthattherearepairsofsearchalgorithmsAandBsuchthat theperformancemeasureused). expectedperformanceofallalgorithmsonthatfunctionareexactlythesame(regardlessof search.inpointoffactthough,asourcentralresultdemonstrates,thisisnotthecase.if wedonottakeintoaccountanyparticularbiasesorpropertiesofourcostfunction,thenthe thisreason(andtoemphasizetheparallelwithsimilarsupervisedlearningresults[16,17]), onlyaswellastheknowledgeconcerningthecostfunctionputintothecostalgorithm.for wehavedubbedourcentralresulta\nofreelunch"(nfl)theorem. Inshort,thereareno\freelunches"foreectiveoptimization;anyalgorithmperforms aspectsofsearch.thisframeworkconstitutesthe\skeleton"oftheoptimizationproblem;it iswhatcanbesaidconcerningsearchbeforeexplicitdetailsofaparticularreal-worldsearch problemareconsidered.theconstructionofsuchaskeletonprovidesalanguagetoaskand ToprovetheNFLtheoremaframeworkhastobedevelopedwhichaddressesthecore nevermindanswered.(weposeandansweranumberofsuchquestionsinthispaper.)in addition,suchaskeletonindicateswherethereal\meat"ofoptimizationlies.itclaries whatthecoreissuesarethatunderlytheeectivenessofthesearchprocess. answerformalquestionsaboutsearch,someofwhichhaveneverbeforeevenbeenasked, andusingittoprovethenfltheorem.weprovethetheoremforbothdeterministicand stochasticsearchalgorithms.section3givesageometricinterpretationofthenfltheorem. Inparticular,inthatsectionweprovideageometricmeaningofwhatitmeansforan Thepaperisorganizedasfollows.Webegininsection2bypresentingourframework 2
3 tigationofthestatisticalnatureofthesearchproblem,usingtheframeworkdevelopedin section2. algorithmtobewell\matched"toacostfunction. Insomecircumstancestheaveragebehaviorofalgorithmsisnotaninterestingquantity TherestofthepapergoesbeyondtheNFLtheorem.Itconsistsofapreliminaryinves- holdforanydistributionovercostfunctions. bywhichtocomparealgorithms.alternatively,averagesmaybeinteresting,butitisn'tclear whatdistributionovercostfunctionstousetodotheaveraging.weaddresssuchscenarios insection4byinvestigatingminimaxdistinctionsbetweenalgorithms.suchdistinctions ofthenfltheoreminanalyzingoptimization.)amyriadofotherpropertiesofsearchmay thatthoseresultsarederivedfromthenfltheorem,theyillustratethecentralimportance answersleadnaturallyintoresultsconcerningtheinformationtheoreticaspectsofsearch.(in Section5beginstheexplorationofsomeofthequestionsraisedinsection2.Someofthe beinvestigatedusingtechniquessimilartothosedevelopedinthissection.welistasample oftheseinsection9.2. ularsearchalgorithms.wederiveseveralbenchmarksagainstwhichtocomparesuchan (ratherthanrelative)ecacyofanalgorithmonsomesearchproblemthatdoesn'tusethese algorithm'sperformance.wecannotconceiveofanyvaliddemonstrationofthe\absolute" InSection6weturntotheimportantproblemofassessingtheperformanceofpartic- (orsimilar)benchmarks. Section7extendsouranalysistothecaseofsuchtimedependentcostfunctions. Notallsearchproblemsarestatic;insomecasesthecostfunctionchangesovertime. onthatfunction.whenchoosingbetweenalgorithmsbasedontheirobservedperformance thereforeforanydistributionovercostfunctions.thesetheoremsstatethatonecannotuse asearchalgorithm'sbehaviorsofaronaparticularcostfunctiontopredictitsfuturebehavior Insection8weprovidesometheoremsvalidforanysinglexedcostfunction,and itdoesnotsucetomakeanassumptionaboutthecostfunction;some(currentlypoorly understood)assumptionsare'alsobeingmadeabouthowthealgorithmsinquestionare relatedtoeachotherandtothecostfunction. results,andthenoffuturedirectionsforwork. Thepapercanbereadinstages.ArstreadingmighthighlighttheNFLtheoremandits Finally,weconcludeinSection9withageneraldiscussionoftheimplicationsofour ofthenfltheorem.finally,section9.1discussesbroadimplicationsofthenflresult. Section4,whichconsidersminimaxdistinctionsbetweenalgorithms,addresseslimitations NFLtheorem,Eq.(1).Section3thenprovidesageometricunderstandingofthetheorem. broadimplications.suchareadingshouldstartwithsection2foranunderstandingofthe Sections2and3.Suchareadingshouldincludesection5,whichusesourframeworkto tosection6whichusestheframeworktoprovideusefulbenchmarksagainstwhichother demonstratesomeoftheinformationtheoreticaspectsofsearch.itwouldthenmoveon Asecondreadingmightexplorethepotentialrichnessoftheframeworkdevelopedin algorithmsmaybecompared. Analreadingwouldincludesubjectsthatmayconstitutefruitfulextensionsofthe 3
4 section8,whichprobeswhatmaybelearnedfromalimitedamountofsearchoverasingle, specic,costfunction.thisreadingwouldconcludewithsection9.2wherewelistmany frameworkdevelopedinsections2and3.suchareadingwouldincludesection7,which extendsthenflresultstoaclassoftime-dependentcostfunctions.itwouldalsoinclude directionsforfutureextensions. sense.ifsomeonewishestocomparealgorithmsonsomeotherbasis,wewishthemluck. numberofdistinctevaluationsofthecostfunctionissimplyourchoice.althoughweconsider itquitereasonable,wedonotclaimtobeableto\prove"thatoneshoulduseit,inany Weshouldemphasizethatourcomparingalgorithmsbasedontheirhavingthesame ofirrelevantaprioridistinctionsbetweenalgorithms.(forexample,itsaysthataglobal totalevaluations includingrepeats isfraughtwithdiculties,andresultsinallkinds Howeverasanasideononesuchcomparisonscheme,wenotethatcomparingbasedon therandomguesserwillretraceless.) randomguesserisbetterthanahill-climber,averagedoverallcostfunctions,simplybecause inparticular,theeldofcomputationalcomplexity.unliketheapproachtakeninthispaper,computationalcomplexityignoresthestatisticalnatureofsearchforthemostpart,and concentratesinsteadoncomputationalissues.much(thoughbynomeansall)ofcomputationalcomplexityisconcernedwithphysicallyunrealizablecomputationaldevices(turing Thereareanumberofotherformalapproachestotheissuesinvestigatedinthispaper, Incontrast,theanalysisinthispaperdoesnotconcernitselfwiththecomputationalengineusedbythesearchalgorithm,butratherconcentratesexclusivelyontheunderlying machines)andtheworstcaseamountofresourcestheyrequiretondoptimalsolutions. (realistic)concernsforcomputationalresources. statisticalnatureofthesearchproblem. Futureworkwouldinvolvecombiningourconcernforthestatisticalnatureofsearchwith associatedcostvalues,(x;y)m2(xy)m,toanewpointx02xthathopefullyhaslowcost 2Alloracle-basedsearchalgorithmsrelyonextrapolatingfromanexistingsetofmpointsand NoFreeLunchTheoremforSearch beeitherdeterministicorstochastic.theanalysisofsuchextrapolationscanbeformalized (highcostifwe'researchingforamaximumratherthanaminimum).theextrapolationmay i=1:::mtobeasetofmdistinctsearchpoints(i.e.costevaluations)andassociatedcost asfollows. valuesorderedinsomeway(usuallyaccordingtothetimeatwhichtheyaregenerated)with theorderingindexgivenbyi.letuscallthisapopulationofsizem.wedenotethesetof ForsimplicitytakeXandYtobenite.Denedmfdm(i)gfdxm(i);dym(i)gfor pointx2xischosenbasedonthemembersofthecurrentpopulationd;thepairfx0;f(x0)g allpopulationsofsizembydm. isaddedtod;andtheprocedurerepeats. areanitenumberoffifjxjandjyjarenite.ateachstageofasearchalgorithm,anew Asabove,letfindicateasingle-valuedfunctionfromXtoY:f2YX.Notethatthere 4
5 hapsprobabilistic)mappingtakinganypopulationtoanewpointinthesearchspace.for simplicityofthepresentation,weassumethatthenewsearchpointhasnotalreadybeen visited.(asdiscussedbelow,relaxingthisassumptiondoesnotaectourresults.)soin Anysearchalgorithmofthe\secondapproach"discussedintheintroductionisa(per- discussedbelow,allourresultsalsoapplytostochasticalgorithms. thispaperwewillonlyexplicitlyconsidersuchdeterministicsearchalgorithms.howeveras D[mDm,andinparticularcontainstheemptyset.Forclarityoftheexposition,in particularadeterministicsearchalgorithmisamappinga:d2d!fxjx62dxg,where tness,itisnecessarytoevaluatethetnessesofalltheneighborsofx.allthoseevaluated pointsarecontainedinthepopulation,notonlyxandtheneighborofxwithhighesttness. ventionalhill-climberthatworksbymovingfromxtothatneighborofxwiththehighest Notethatthepopulationcontainsallpointssampledsofar.Inparticular,inacon- particularcostfunction,f,givenmdistinctcostevaluations.notethat~cisgivenbythey valuesofthepopulation,dym,andisavectoroflengthjyjwhoseithcomponentisthenumber ofmembersinthepopulationdmhavingcostfi.oncewehave~cwecanuseittoassessthe Weareinterestedinthehistogram,~c,ofcostvaluesthatanalgorithm,a,obtainsona ofalgorithmaonf.thisquantityisgivenbytheconditionalprobabilityp(~cjf;m;a). interestedintheconditionalprobabilitythathistogram~cwillbeobtainedundermiterations mighttakethelowestoccupiedbinin~casourperformancemeasure.)consequently,weare qualityofthesearchinanywaywechoose.(forexampleifwearesearchingforminimawe allfofp(~cjf;m;a1)tothesumoverallfofp(~cjf;m;a1).thiscomparisonprovidesa reverseistrue.toperformthecomparison,weusethetrickofcomparingthesumover algorithma1outperformsanotheralgorithma2,comparestof2,thesetoffforwhichthe AnaturalquestionconcerningthisscenarioishowF1,thesetoffforwhichsome majorresultofthispaper:p(~cjf;m;a)isindependentofawhenweaverageoverallcost functions.inotherwords,asisprovenbelow, Theorem:Foranypairofalgorithmsa1anda2, Animmediatecorollaryisthatforanyperformancemeasure(~c),theaverageoverallf XfP(~cjf;m;a1)=XfP(~cjf;m;a2): (1) toaperformancemeasureisirrelevant. ofp((~c)jf;m;a)isindependentofa.sotheprecisewaythatthehistogramismapped ofalgorithma,isindependentofa.thisfollowsfrom P(~cjm;a),whichistheprobabilityweobtainhistogramcaftermdistinctcostevaluations Notethatthenofreelunchresultimpliesthatifweknownothingaboutf,then (inthelaststepwehavereliedonthefactthatthecostfunctiondoesn'tdependoneither P(~cjm;a)=XfP(~cjf;m;a)P(fjm;a)=XfP(~cjf;m;a)P(f) mora).ifweknownothingaboutfthenallfareequallylikely,whichmeansthatfor allf,p(f)=1=jyjjxj.(moregenerally,p(f)reectsour\priorknowledge"concerningf.) 5
6 isindependentofabythenofreelunchtheorem. Accordingly,forthis\noknowledge"scenario,P(~cjm;a)=jYj jxjpfp(cjf;m;a),which ofthespace.ratheritisthetypicalcase. possiblep(f).)inthis,theuniformp(f)caseisnotsome\pathologicalcase",ontheedge theresultconcernsaveragingoverallthequantityp(~cjm;),whereindexesthesetof Similarly,youcanderiveanNFLresultforaveragingoverallpriors.(Moreformally, ifalgorithma1hasbetterperformancethanalgorithma2oversomesubsetfoffunctions,thena2mustperformbetteronthesetofremainingfunctionsfn.soforexampleif maximumofthecostfunction,hill-climbingandhill-descendingareequivalent,onaverage. simulatedannealingoutperformsgeneticalgorithmsonsomeset,geneticalgorithmsmust outperformsimulatedannealingonfn.asanotherexample,evenifone'sgoalistonda P~c~cP(~cjf;m;a)is,onaverage,thesameforallalgorithms.Moregenerally,foranytwo algorithms,atthepointintheirsearchwheretheyhavebothcreatedapopulationofsizem, AnotherimmediateconsequenceoftheNFLresultisthattheexpectedhistogramE(~cjf;m;a)= Thereareasmanyfforwhichyouralgorithm'sguessesforwheretosearchareworsethan ofrandomsearch.thenflresultsaysthatthereareasmanyf(appropriatelyweighted) forwhichtherandomalgorithmoutperformsyourfavoritesearchalgorithmasvice-versa. Aparticularlystrikingexampleofthislastpointisthecasewherea2isthealgorithm itmayperformrandomlyonthefathand,butthatitmayverywellperformevenworse. randomasforwhichtheyarebetter.theriskyoutakeinchoosinganalgorithmisnotthat somethingaboutf(perhapsspeciedthroughp(f)),ifwefailtoexplicitlyincorporatethat ofthisisdemonstratedbythenfltheorem,whichillustratesthatevenifwedoknow veryrarelyisthatknowledgeexplicitlyusedtohelpsetthealgorithm.theunreasonableness Oftenintherealworldonehassomeaprioriknowledgeconcerningf.Howeveronly onafortuitousmatchingbetweenfanda.thispointisformallyestablishedinsections3 knowledgeintoathenwehavenoassurancestheawillbeeective;wearesimplyrelying obvious.similarly,itmayseemobviousthatifoneuniformlyaveragesoverallf,thenall and8,whichmakenoassumptionswhatsoeverconcerningp(f). algorithmsareequal.(theonlyreasonittakesawholesubsectiontoestablishthisformally isbecausetherearealargenumberof\obvious"thingsthatmustbemathematicized.)yet ManywouldreadilyagreethatamustmatchP(f) thatstatementbordersonthe climbingandhill-descendingareequivalentonaverage,orthat\smart"choosingprocedures performnobetterthan\dumb"ones(seesection8).inaddition,thegeometricnatureof withoutrealizingyouaredoingso.thisiswhy,forexample,itcanbesurprisingthathill- theimplicationsofthestatementarenotsoobvious;itisextremelyeasytocontradictthem search.itistheonlystartingpointwecouldthinkofforinvestigatingthe\skeleton"ofthe searchproblem,before(assumptionsfor)theactualdistributionsintherealworldareput thematchingillustratessomeinterestingaspectsofthesearchproblem(seebelow). in.itshouldbeobviousthatwearenotclaimingthatallf'sareequallylikelyinthereal Weemphasizethattakinguniformaveragesoverf'sissimplyatoolforinvestigating world,andthesignicanceofthenfltheoreminnowaydependsonthevalidityofsucha claim. Resultsfornon-uniformP(f)arediscussedbelow,aftertheproofoftheNFLtheorem. 6
7 WenowshowthatPfP(~cjf;m;a)hasnodependenceona.Conceptually,theproofis quitesimple;theonlyreasonittakessolongisbecausethereissomebook-keepinginvolved. 2.1 Prooffordeterministicsearch Inaddition,becausemanyofourreadersmaynotbeconversantwiththetechniquesof probabilitytheorywesupplyallthedetails,lengtheningitconsiderably. values.thenweuseinductiontoestablishthea-independenceofthedistributionoverdym. involvesthefollowingsteps:first,wereducethedistributionover~cvaluestooneoverdym hasnobearingonitsfutureperformancesothatallalgorithmsperformequally.theproof Theintuitionissimple:bysummingoverallfthepastperformanceofanalgorithm separately,givingthedesiredresult. upintotwoindependentparts,oneforx2dxmandoneforx62dxm.theseareevaluated Theinductivestepstartsbyrearrangingthedistributionsinquestion.Thenfisbroken Expandingoverallpossibleycomponentsofapopulationofsizem,dym,wesee NowP(~c;dymjf;m)=P(~cjdym;f;m;a)P(dymjf;m;a).Moreover,theprobabilityofobtainingahistogram~cgivenf,d,manda,P(~cjdym;f;m),dependsonlyontheyvaluesof populationdm.therefore XfP(~cjf;m;a)=Xf;dymP(~cjdym)P(dymjf;m;a) XfP(~cjf;m;a)=Xf;dymP(~c;dymjf;m;a) ToprovethattheexpressioninEq.(2)isindependentofaitsucestoshowthatfor =XdymP(~cjdym)XfP(dymjf;m;a) allmanddym,pfp(dymjf;m;a)isindependentofa,sincep(~cjdym)isindependentofa.we willprovethisbyinductiononm. possiblevaluefordy1isf(dx1),sowehave: Form=1wewritethepopulationasd1=fdx1;f(dx1)gwheredx1issetbya.Theonly whereisthekroneckerdeltafunction. XfP(dy1jf;m=1;a)=Xf(dy1;f(dx1)) whichhavecostdy1atpointdx1.thereforethatsumequalsjyjjxj 1,independentofdx1: Nowwhenwesumoverallpossiblecostfunctions(dy1;f(dx1))is1onlyforthosefunctions whichisindependentofa.thisbasestheinduction. XfP(dy1jf;m=1;a)=jYjjXj 1 dym,thensoalsoispfp(dym+1jf;m+1;a).thiswillcompletetheproofofthenflresult. Wenowestablishtheinductivestep,thatifPfP(dymjf;m;a)isindependentofaforall 7
8 Westartbywriting P(dym+1jf;m+1;a)=P(fdym+1(1);:::;dym+1(m)g;dym+1(m+1)jf;m+1;a) sowehave =P(dym;dym+1(m+1)jf;m+1;a) XfP(dym+1jf;m+1;a)=XfP(dym+1(m+1)jdym;f;m+1;a)P(dymjf;m+1;a): =P(dym+1(m+1)jdm;f;m+1;a)P(dymjf;m+1;a) weexpandoverthesepossiblexvalues,getting Thenewyvalue,dym+1(m+1),willdependonthenewxvalue,fandnothingelse.So XfP(dym+1jf;m+1;a)=Xf;xP(dym+1(m+1)jf;x)P(xjdym;f;m+1;a) =Xf;x(dym+1(m+1);f(x))P(xjdym;f;m+1;a) P(dymjf;m+1;a) expandindxmtoremovethefdependenceinp(xjdym;f;m+1;a): Nextnotethatsincex=a(dxm;dym),itdoesnotdependdirectlyonf.Consequentlywe P(dymjf;m+1;a): XfP(dym+1jf;m+1;a)=X =Xf;dxm(dym+1(m+1);f(a(dm)))P(dmjf;m;a) f;x;dxm(dym+1(m+1);f(x))p(xjdm;a)p(dxmjdym;f;m+1;a) P(dymjf;m+1;a) whereusewasmadeofthefactthatp(xjdm;a)=(x;a(dm))andthefactthatp(dmjf;m+ 1;a)=P(dmjf;m;a). pointsrestrictedtodxmandthosepointsoutsideofdxm.p(dmjf;m;a)willdependonthef valuesdenedoverpointsoutsidedxm.(recallthata(dxm)62dxm.)sowehave valuesdenedoverpointsinsidedxmwhile(dym+1(m+1);f(a(dm)))dependsonlyonthef Wedothesumovercostfunctionsfrst.Thecostfunctionisdenedbothoverthose XfP(dym+1jf;m+1;a)=XdxmX X f(x2dxm) (dym+1(m+1);f(a(dm))): P(dmjf;m;a) ThesumPf(x62dxm)contributesaconstant,jYjjXj m 1,equaltothenumberoffunctions (3) denedoverpointsnotindxmpassingthrough(dxm+1(m+1);f(a(dm))).so XfP(dym+1jf;m+1;a)=jYjjXj m 1f(x2dxm);dxm X 8 P(dmjf;m;a)
9 = jyjxf;dxmp(dmjf;m;a) Byhypothesistherighthandsideofthisequationisindependentofa,sothelefthandside jyjxfp(dymjf;m;a) 1 mustalsobe.thiscompletestheproofofthenflresult. ofcostvaluesaftermstepsmustalsobeindependentofa.however,italsofollowsthat result.sincethesumpfp(dymjf;m;a)isindependentofa,itfollowsthatthehistograms thedistributionovertimeorderedpopulations(thedym)arealsoidenticalforalla.sowhen WenoteinpassingthattheproofoftheNFLtheoremcanbeusedtoderiveastronger theorderingofcostvaluesisimportant(e.gwhenyouwouldliketogettolowcostquickly) thereisstillnowaytodistinguishbetweenalgorithmswhenweaverageoverallf. ndobjectionable.theseare:i)thebanningofalgorithmsthatmightrevisitthesamepoints 2.2 Therearetworestrictionsonthedenitionofsearchalgorithmsusedsofarthatonemight Moregeneralkindsofsearch eitheralgorithmsthatrevisitpointsand/orarealgorithmsthatarestochastic.sothereis ratherthandeterministically.fortunately,thenflresultcaneasilybeextendedtoinclude nolossofgeneralityinourdenitionofa\searchalgorithm". inxafterplacingthemindx;andii)thebanningofalgorithmsthatworkstochastically algorithma0by\skippingoverallduplications"inthesequenceoffx;ygpairsproduced algorithm\potentiallyretracing".givenapotentiallyretracingalgorithma,produceanew givensome(perhapsempty)d,thealgorithmmightproduceapointx2dx.callsuchan Toseethis,saywehaveadeterministicalgorithma:d2D!fxjx2Xg,sothat originalalgorithmacannotgetstuckforeverinsomesubsetofd,wecanalwaysproduce bythepotentiallyretracingalgorithm.formally,foranyd,a0(d)isdenedastherstx valuefromthesequencefa(;);a(d);a(a(d));:::gthatisnotcontainedindx.solongasthe suchana0froma.(wecanndnoreasontodesignone'salgorithmtonothavean\escape thata0isa\compacted"versionofa. mechanism"thatensuresthatitcannotgetstuckforeverinsomesubsetofd.)wewillsay intheprevioussubsection.thereforetheyobeythenflresultofthatsubsection.sothe thatequationtobethenumberofdistinctpointsinthedx'sproducedbythealgorithms,in NFLresultinEq.(1)holdsevenforpotentiallyretracingalgorithms,ifweredene`m'in Nowanytwocompactedalgorithmsare\searchalgorithms"inthesensethetermisused question,andifweredene`~c'tobethehistogramcorrespondingtothosemdistinctpoints. bylookingatthed'stheyproduceaftersamplingf(x)thesamenumberoftimes.thisis distinctevaluationsoff(x).soitmakessensetocomparepotentiallyretracingalgorithms notbylookingatthed'stheyproduceafterbeingrunthesamenumberoftimes,butrather Moreover,ourreal-worldcostinusinganalgorithmisusuallysetbythenumberof consistentwithusingourredenedmand~c. 9
10 isstillwell-dened.onlyratherthanbeingdeterministic,thatcompactedalgorithmis stochastic.thisbringsustothegeneralissueofhowtoadaptouranalysistoaddress bestochastic(e.gsimulatedannealing).inthiscasethecompactedversionofthealgorithm Notethatthexatwhichapotentiallyretracingalgorithmbreaksoutofacyclemight stochasticsearchalgorithms. amappingtakinganydtoa(d-dependent)distributionoverxthatequalszeroforallx2dx. Socanbeviewedasa\hyper-parameter",specifyingthefunctionP(dxm+1(m+1)jdm;) forallmandd. Letbeastochasticnon-potentiallyretractingalgorithm.Formally,thismeansthatis stillholds.sothatnflresultholdsevenforstochasticsearchalgorithms.therefore, bythesamereasoningusedtoestablishtheno-free-lunchresultforpotentiallyretracing fordeterministicalgorithms,justwithareplacedbythroughout.doingso,everything Giventhisdenitionof,wecanfollowalongwiththederivationoftheNFLresult deterministicalgorithms,theno-free-lunchresultholdsforpotentiallyretracingstochastic algorithms. speciedthroughp(f))butdon'tincorporatethatknowledgeintoa,thenwehavenoassurancesthatawillbeeective;wearesimplyrelyingonafortuitousmatchingbetween Intuitively,theNFLtheoremillustratesthatevenifweknowsomethingaboutf(perhaps 3 Ageometricinterpretation fanda.thispointisformallyestablishedbyviewingthenfltheoremfromageometric perspective. obtainingsomehistogram,~c,givenmdistinctcostevaluationsusingalgorithmais Considerthespaceofpossiblecostfunctions.Asmentionedbefore,theprobabilityof wherep(f)isthepriorprobabilitythattheoptimizationproblemathandhascostfunction P(~cjm;a)=XfP(~cjm;a;f)P(f): f.wecanviewtheright-handsideofthisequalityasaninnerproductinf: Theorem:DenetheF-spacevectors~vc;a;mand~pby~vc;a;m(f)P(~cjm;a;f)and~p(f) P(f).Then yourcostfunctiongoesintotheprior,~p,overcostfunctions.~ccanbeviewedasxedto Thisisanimportantequation.Anyglobalknowledgeyouhaveaboutthepropertiesof P(~cjm;a)=~vc;a;m~p (4) theconstraintsonthetimewehavetorunouroptimizationalgorithm.thustheoptimal thehistogramyouwanttoobtain(usuallyonewithalowcostvalue),andmisgivenby algorithmisthatwhichhasthelargestprojectiononto~p.alternatively,wecandispense 10
11 P(f)must\match"a. E(~cjm;a;f).(Similarlyforany\performancemeasure"(~c).Ineithercase,weseethat with~cbyaveragingoverit,toseethate(~cjm;a)isaninnerproductbetween~p(f)and P(f)canbedicult.Consider,forexample,doingTSPproblemswithNcities.Sowe're onlyconsideringcostfunctionsthatcorrespondtosuchaproblem.nowtothedegree thatanypractitionerwouldattackalln-citytspcostfunctionswiththesamealgorithm, Ofcourse,exploitingthisinpracticeisadicultexercise.Evenwritingdownareasonable thatpractitionerimplicitlyignoresdistinctionsbetweensuchcostfunctions.inthis,that practitionerhasimplicitlyagreedthattheproblemisoneofhowtheirxedalgorithmdoes acrossthesetofn-citytspcostfunctions,ratherthanofhowwelltheiralgorithmdoesfor thefactthatitisrestrictedton-citytspproblems,maybeverydiculttodisentangle. thoughthecostfunctionwerenotxed,butisinsteaddescribedbyap(f)thatequals0for allcostfunctionsotherthann-citytspcostfunctions.howeverthedetailsofp(f),beyond someparticularn-citytspproblemtheyhaveathand.inotherwords,theyareactingas ofahasthesimpleinterpretationthatforaparticular~candm,allalgorithmsahavethe sameprojectionontothediagonal,thatisvc;a;m~1=cst(~c;m).fordeterministicalgorithms thecomponentsofvc;a;m(i.e.,theprobabilitiesthatalgorithmagiveshistogram~concost Takingthegeometricview,thenofreelunchresultthatPfP(~cjf;m;a)isindependent alsoimpliespfp2(~cjm;a;f)=cst(~c;m).geometrically,thismeansthatthelengthof~vc;a;m isindependentofa. functionfaftermdistinctcostevaluations)arealleither0or1sothenofreelunchresult thesubsetofthebooleanhypercubehavingthesamehammingdistancefrom~0. onto~1.becausethecomponentsof~vc;a;marebinarywemightalsoview~vc;a;maslyingon Thusallvectors~vc;a;mhavethesamelengthandlieonaconewithconstantprojection ~c.thisisingeneralanjfj 2dimensionalmanifold(wherewerecallthatjFjjYjjXjis particular~c.thealgorithmsinthissetmustlieintheintersectionof2cones oneabout thediagonal,setbytheno-free-lunchtheorem,andonebyhavingthesameprobabilityfor Nowrestrictattentiontothesetofalgorithmsthathavethesameprobabilityofsome thenumberofpossiblecostfunctions).ifwerequireequalityofprobabilityonyetmore~c, wegetyetmoreconstraints. ofthishypercube. InSection5wecalculatetwoquantitiesconcerningthedistributionof~vc;a;macrossvertices TheNFLtheoremdoesnotaddressminimaxpropertiesofsearch.Forexample,saywe're consideringtwodeterministicalgorithms,a1anda2.itmayverywellbethatthereexist 4 Minimaxdistinctionsbetweenalgorithms costfunctionsfsuchthata1'shistogramismuchbetter(accordingtosomeappropriate qualitymeasure)thana2's,butnocostfunctionsforwhichthereverseistrue.forthe NFLtheoremtobeobeyedinsuchascenario,itwouldhavetobetruethattherearemany betterforallthosef.forsuchascenario,inacertainsensea1hasbetter\head-to-head" morefforwhicha2'shistogramisbetterthana1'sthanvice-versa,butitisonlyslightly 11
12 minimaxbehaviorthana2;therearefforwhicha1beatsa2badly,butnoneforwhicha1 doessubstantiallyworsethana2. denitioncanbeusedifoneisinsteadinterestedin(~c)ordymratherthan~c.) gorithmsa1anda2ithereexistsaksuchthatforatleastonefe(~cjf;m;a1) E(~cj f;m;a2)=k,butthereisnofsuchthate(~cjf;m;a2) E(~cjf;m;a1)=k.(Asimilar Formally,wesaythatthereexistshead-to-headminimaxdistinctionsbetweentwoal- moredicultthananalyzingaveragebehavior(likeinthenfltheorem).presently,very littleisknownaboutminimaxbehaviorinvolvingstochasticalgorithms.inparticular,itis notknownifinsomesenseastochasticversionofadeterministicalgorithmhasbetter/worse Itappearsthatanalyzinghead-to-headminimaxpropertiesofalgorithmsissubstantially todeterministicalgorithms,onlyanextremelypreliminaryunderstandingofminimaxissues hasbeenreached. minimaxbehaviorthanthatdeterministicalgorithm.infact,evenifwestickcompletely Whatwedoknowisthefollowing.Considerthequantity fordeterministicalgorithmsa1anda2(bypa(a)ismeantthedistributionofarandom XfPdym;1;dym;2(z;z0jf;m;a1;a2); andthata2producesapopulationwithycomponentsz0. numberoffsuchthatitisbothtruethata1producesapopulationwithycomponentsz variableaevaluatedata=a).fordeterministicalgorithms,thisquantityisjustthe Theorem:Ingeneral, interchangeofzandz0: InappendixB,itisprovenbyexamplethatthisquantityneednotbesymmetricunder Thismeansthatundercertaincircumstances,evenknowingonlytheYcomponentsofthe XfPdym;1;dym;2(z;z0jf;m;a1;a2)6=XfPdym;1;dym;2(z0;zjf;m;a1;a2): (5) thingconcerningwhatalgorithmproducedeachpopulation. populationsproducedbytwoalgorithmsrunonthesame(unknown)f,wecaninfersome- NowconsiderthequantityXfPC1;C2(z;z0jf;m;a1;a2); againfordeterministicalgorithmsa1anda2.thisquantityisjustthenumberoffsuchthat itisbothtruethata1producesahistogramzandthata2producesahistogramz0.ittoo statementthentheasymmetryofdy'sstatement,sinceanyparticularhistogramcorresponds tomultiplepopulations. neednotbesymmetricunderinterchangeofzandz0(seeappendixb).thisisastronger a1anda2suchthatforsomefa1'shistogramismuchbetterthana2's,butfornof'sisthe reverseistrue.toinvestigatethisprobleminvolveslookingoverallpairsofhistograms(one Itwouldseemthatneitherofthesetworesultsdirectlyimpliesthattherearealgorithms 12
13 foreachf)suchthatthereisthesamerelative\quality"betweenbothhistograms.simply havinganinequalitybetweenthesumspresentedabovedoesnotseemtodirectlyimplythat therelativequalitybetweentheassociatedpairofhistogramsisasymmetric.(toformally establishthiswouldinvolvecreatingscenariosinwhichthereisaninequalitybetweenthe sums,butnohead-to-headminimaxdistinctions.suchananalysisisbeyondthescopeof thispaper.) forallothers.insuchacase,pfpdym;1;dym;2(z1;z2jf;m;a1;a2)isjustthenumberoffthatresultinthepair(z1;z2).sopfpdym;1;dym;2(z;z0jf;m;a1;a2)=pfpdym;1;dym;2(z0;zjf;m;a1;a2tic,thenforanyparticularfpdym;1;dym;2(z1;z2jf;m;a1;a2)equals1forone(z1;z2)pair,and0 therearehead-to-headminimaxdistinctions.forexample,ifbothalgorithmsaredeterminis- Ontheotherhand,havingthesumsequaldoescarryobviousimplicationsforwhether impliesthattherearenohead-to-headminimaxdistinctionsbetweena1anda2.theconverse doesnotappeartoholdhowever.1 denethefollowingmeasureofthe\quality"overtwo-elementpopulations,q(dy2): canexploittheresultinappendixb,whichconcernsthecasewherejxj=jyj=3.first, Asapreliminaryanalysisofwhethertherecanbehead-to-headminimaxdistinctions,we ii)q(y1;y2)=q(y2;y1)=0. i)q(y2;y3)=q(y3;y2)=2. iii)qofanyotherargument=1. histogramfy2;y3ganda2generatesfy1;y2g). thatforonefa1generatesthehistogramfy1;y2ganda2generatesthehistogramfy2;y3g, butthereisnofforwhichthereverseoccurs(i.e.,thereisnofsuchthata1generatesthe InappendixBweshowthatforthisscenariothereexistpairsofalgorithmsa1anda2such otherfforwhichthedierenceis-2.forthisqthen,algorithma2isminimaxsuperiorto ThedierenceintheQvaluesforthetwoalgorithmsis2forthatf.Howeverthereareno betweena1anda2.foronefthequalityofalgorithmsa1anda2arerespectively0and2. Sointhisscenario,withourdenedmeasureof\quality",thereareminimaxdistinctions algorithma1. maxifdym(i)gtherearenominimaxdistinctionsbetweenalgorithms. distinctionsbetweenthealgorithms.asanexample,itmaywellbethatforq(dym)= Moregenerally,atpresentnothingisknownabout\howbigaproblem"thesekindsof ItisnotcurrentlyknownwhatrestrictionsonQ(dym)areneededfortheretobeminimax asymmetriesare.alloftheexamplesoftheasymmetriesarisewhenthesetofxvaluesa1 lunchtheorem,thesumofallnumbersinrowzequalsthesumofallnumbersincolumnz.thesetwo point's(z;z0)pair.thenourconstraintsarei)bythehypothesisthattherearenohead-to-headminimax distinctions,ifgridpoint(z1;z2)isassignedanon-zeronumber,thensois(z2;z1);andii)bytheno-free- 1Considerthegridofall(z;z0)pairs.Assigntoeachgridpointthenumberoffthatresultinthatgrid andcolumns.althoughagain,likebefore,toformallyestablishthispointwouldinvolveexplicitlycreating constraintsdonotappeartoimplythatthedistributionofnumbersissymmetricunderinterchangeofrows searchscenariosinwhichitholds. 13
14 those\certainproperties"isnotyetinhand.norisitknownhowgenerictheyare,i.e.,for ofhowthealgorithmsgeneratedtheoverlap,asymmetryarises.aprecisespecicationof whatpercentageofpairsofalgorithmstheyarise.althoughsuchissuesareeasytostate hasvisitedoverlapswiththosethata2hasvisited.givensuchoverlap,andcertainproperties (seeappendixb),itisnotatallclearhowbesttoanswerthem. donotoverlap.suchassuranceshold,forexample,ifwearecomparingtwohill-climbing assurances,therearenoasymmetriesbetweenthetwoalgorithmsform-elementpopulations. algorithmsthatstartfarapart(onthescaleofm)inx.itturnsoutthatgivensuch Howeverconsiderthecasewhereweareassuredthatinmstepstwoparticularalgorithms Doingthisestablishesthefollowing: thoseargumentstothequantitypfpdym;1;dym;2(z;z0jf;m;a1;a2)ratherthanp(~cjf;m;a). Toseethisformally,gothroughtheargumentusedtoprovetheNFLtheorem,butapply Theorem:Ifthereisnooverlapbetweendxm;1anddxm;2,then Animmediateconsequenceofthistheoremisthatundertheno-overlapconditions,PfPC1;C2(z;z0j XfPdym;1;dym;2(z;z0jf;m;a1;a2)=XfPdym;1;dym;2(z0;zjf;m;a1;a2): (6) f;m;a1;a2)issymmetricunderinterchangeofzandz0,asarealldistributionsdetermined isalwaysoverlaptoconsider.sothereisalwaysthepossibilityofasymmetrybetween extrema). fromthisoneoverc1andc2(e.g.,thedistributionoverthedierencebetweenthosec's algorithmsifoneofthemisstochastic. Notethatwithstochasticalgorithms,iftheygivenon-zeroprobabilitytoalldxm,there Werstcalculatethefractionofcostfunctionswhichgiverisetoaspecichistogram~cusing 5algorithmawithmdistinctcostpoints.Thiscalculationallowsus,forexample,toanswer Informationtheoreticaspectsofsearch thefollowingquestion: distinctcostevaluationschosenbyusingageneticalgorithm?" \Whatfractionofcostfunctionswillgiveaparticulardistributionofcostvaluesafterm thisbecauseitmeansthatthefractionisindependentofthealgorithm!sowecananswer thequestionbyusinganalgorithmforwhichthecalculationisparticularlyeasy. Thismayseemanintractablequestion,buttheNFLresultallowsustoanswerit.Itdoes x1;x2;:::;xm.recallthatthehistogram~cisspeciedbygivingthefrequenciesofoccurrence, acrossthex1;x2;:::;xm,foreachofthejyjpossiblecostvalues. ThealgorithmwewilluseisonewhichvisitspointsinXinsomecanonicalorder,say justthemultinomialgivingthenumberofwaysofdistributingthecostvaluesin~c.atthe remainingjxj mpointsinxthecostcanassumeanyofthejyjfvalues. Nowthenumberoff'sgivingthedesiredhistogramunderourspeciedalgorithmis 14
15 binsin~carescaledbythesameamount.bytheargumentoftheprecedingparagraph,the fractionweareinterestedin,f(~),isgivenbythefollowing: Itwillbeconvenienttodene~1m~c.Notethatthisisinvariantifthecontentsofall ~c=m~isgivenby Theorem:Foranyalgorithm,thefractionofcostfunctionsthatresultinthehistogram f(~)=c1c2cjyjjyjjxj m mjyjjxj =c1c2cjyj jyjm m Stirling'sapproximationtoorderO(1=m),whichisvalidwhenalloftheciarelarge: Accordingly,f(~)canberelatedtotheentropyof~cinthestandardwaybyusing : (7) ln c1c2cjyj!=mlnm jyj m =ms(~)+12h(1 jyj)lnm jyj Xi=1cilnci+12hlnm jyj Xi=1lnii Xi=1lncii wheres(~)= PjYj thefractionofcostfunctionsisgivenbythefollowingformula: Corollary: i=1ilniistheentropyofthehistogram~c.thusforlargeenoughm, wherec(m;jyj)isaconstantdependingonlyonmandjyj. f(~)=c(m;jyj)ems(~) QjYj i=11=2 i: (8) Eq.(8)canbefoundbysummingoverall~lyingontheunitsimplex.Thedetailsofsuch correspondingtothezero-valued~i.howeveryisdened,thenormalizationconstantof acalculationcanbefoundin[15]. Ifsomeofthe~iare0,Eq.(8)stillholds,onlywithYredenedtoexcludethey's algorithmsthatgiverisetoaparticular~c?" \Onagivenvertexoff-space(i.e.,foragivencostfunction),whatisthefractionofall Wenextturntoarelatedquestion: allx)ofcostvalues.specifythishistogramby~;thereareni=ijxjpointsinxfor whichf(x)hasthei'thyvalue. Forthisquestion,theonlysalientfeatureoffisitshistogram(formedbylookingacross withthefollowingintuitivelyreasonableresult,formallyproveninappendixa: leadingorderonthekullback-liebler\distance"[3]between~and~.toseethis,westart Callthefractionweareinterestedinalg(~;~).Itturnsoutthatalg(~;~)dependsto 15
16 toahistogram~c=m~isgivenby Theorem:Foragivenfwithhistogram~N=jXj~,thefractionofalgorithmsthatgiverise alg(~;~)=qjyj i=1ni jxj m: ci costvaluesfromx.2 Thenormalizationfactorinthedenominatorissimplythenumberofwaysofselectingm (9) ciarelarge: TheproductofbinomialscanbeapproximatedviaStirling'sequationwhenbothNiand lnjyj Yi=1 ci!=jyj Xi=1 12ln2+NilnNi cilnci (Ni ci)ln(ni ci)+ z z2=2 :::,tosecondorderinci=niwehave Weassumeci=Ni1,whichisreasonablewhenmjXj.Sousingtheexpansionln(1 z)= 12(lnNi ln(ni ci) lnci): lnjyj Yi=1 ci!=jyj Xi=1ciln(Ni ci 2Nici 1+ ci) 12lnci+ci 12ln2 Intermsof~and~wenallyobtain(usingm=jXj1) lnjyj Yi=1 Ni ci!= mdkl(~;~)+m mln(m jyj Xi=112ln(im)+m 2jXj(i jxj) jyj i)(1 im+); 2ln2 wheredkl(~;~)piiln(i=i)isthekullback-lieblerdistancebetweenthedistributions Corollary: ~and~. Thusthefractionofalgorithmsisgivenbythefollowing: wheretheconstantcdependsonlyonm,jxj,andjyj. alg(~;~)=c(m;jxj;jyj)e mdkl(~;~) Asbefore,Ccanbecalculatedbysumming~overtheunitsimplex. QjYj i=11=2 i : (10) 2ItcanalsobedeterminedfromtheidentityP~c(Pici;m)Qi Ni 16 ci= PiNi m.
17 Inthissectionwecalculatecertain\benchmark"performancemeasuresthatallowusto assesstheecacyofanysearchalgorithm. 6 Measuresofalgorithmperformance interestedinp(min(~c)>jf;m;a),whichistheprobabilitythattheminimumcostan f.weconsiderthreequantitiesthatarerelatedtothisconditionalprobabilitythatcanbe algorithmandsinmdistinctevaluationsislargerthan,giventhatthecostfunctionis Considerthecasewherelowcostispreferabletohighcost.Theningeneralweare usedtogaugeanalgorithm'sperformance: ii)thesecondistheformthisconditionalprobabilitytakesfortherandomalgorithm, i)therstquantityistheaverageofthisprobabilityoverallcostfunctions. iii)thethirdisthefractionofalgorithmswhich,foraparticularfandm,resultina~c whoseminimumexceeds. whosebehaviorisuncorrelatedwiththecostfunction. jobṙecallthattherearejyjdistinctcostvalues.withnolossofgeneralityassumethei'th whenusedintherealworld;anyalgorithmthatdoesn'tsurpassthemisdoingaverypoor Thesemeasuresgiveusbenchmarkswhichalltruly\intelligent"algorithmsshouldsurpass increments. costvaluesequalsi.socostvaluesrunfromaminimumof1toamaximumofjyjininteger Therstofourbenchmarkmeasuresis PfP(min(~c)>jf;m;a) Pf1 =Pdym;fP(min(dym)>jdym)P(dymjf;m;a) thatmin(c)=min(dym). whereinthelastlinewehavemarginalizedoveryvaluesofpopulationsofsizemandnoted jyjjxj (11) a.inparticular,itequals1ifthefollowingconditionsaremet i)f(dxm(1))=dym(1) NowconsiderPfP(dymjf;m;a).Thesummandequals0or1forallfanddeterministic iii)f(a[dm(1);dm(2)])=dym(3) ii)f(a[dm(1)])=dym(2) atallotherpoints.thereforexfp(dymjf;m;a)=jyjjxj m: Theserestrictionswillalwaysxthevalueoff(x)atexactlympoints.fiscompletelyfree ::: 17
18 UsingthisresultinEq.(11)wend XfP(min(~c)>jf;m)= jyjmxdymp((min(dym)>jdym) = jyjm(jyj )m: 1dym3min(dym)>1 X Theorem: Thisestablishesthefollowing: where!()1 =jyjisthefractionofcostlyingabove. XfP(min(~c)>jf;m)=!m(): (12) Corollary:InthelimitofjYj!1, Animmediatecorrolaryisthefollowing: PfE(min(~c)jf;m) Proofsketch:WritePfE(min(~c)jf;m)=PjYj jyj = m+1: 1 =1[!m( 1)!m()]andsubstituteinfor (13)!().Thenreplacethroughoutwith+1.ThisturnsoursumintoPjYj 1 usethefactthatisgoingto0tocanceltermsinthesummand.carryingthroughthe by.totakethelimitof!0,applyl'hopital'sruletotheratiointhesummand.next!yj)m (1 +1!Yj)m].Next,writejYj=b=forsomeb.Multiplyanddivideoursummand =0[+1][(1 algebra,anddividingbyb=,wegetariemannsumoftheformmb2rb0dxx(1 x=b)m 1. Evaluatingtheintegralgivestheresultclaimed.QED. randomlychosencostfunction.(benchmarksthattakeaccountoftheactualcostfunction thedropassociatedwiththeseresults,onemightarguethatthatalgorithmisnotsearching verywell.afterall,thealgorithmisdoingnobetterthanonewouldexpectittofora Inarealworldscenario,unlessone'salgorithmhasitsbest-cost-so-fardropfasterthan athandarepresentedbelow.) informationfromthecurrentpopulation.marginalizingoverhistograms~c,theperformance ofmfortherandomalgorithm,~a,whichpickspointsinxcompletelyrandomly,usingno of~ais Nextwecalculatetheexpectedminimumofthecostvaluesinthepopulationasafunction P(min(~c)jf;m;~a)=X~cP(min(~c)j~c)P(~cjf;m;~a) 18
19 hasbeencalculatedpreviouslyasqjyj histogram~nofthefunctionf.(thiscanbeviewedasthedenitionof~a.)thisprobability NowP(~cjf;m;~a)istheprobabilityofobtaininghistogram~cinmrandomdrawsfromthe (jxj i=1(ni m)).so ci) P(min(~c)jf;m;~a)= jxj mmx 1c1=0mX cjyj=0(jyj jyj Xi=1ci;m)P(min(~c)j~c) Yi=1 Ni ci! = =PjYj jxj mmxc=0mx 1 cjyj=0(jyj Xi=ci;m)jYj Yi= Ni ci! jxj i=ni ()jxj jxj m m (seefootnoteone) Theorem:Fortherandomalgorithm~a, Thisestablishesthefollowing: (14) where()pjyj i=ni=jxjisthefractionofpointsinxforwhichf(x). P(min(~c)jf;m;~a)=m 1 Yi=0() i=jxj 1 i=jxj: (15) Corollary: Torstorderin1=jXjthistheoremgivesthefollowingresult: Notethattheseresultsallowustocalculateotherquantitiesofinterest,like P(min(c)>jf;m;~a)=m()1 m(m 1)(1 ()) 2() jxj+:::: 1 (16) E(min(~c)jf;m;~a)= Theseresultsalsoprovideausefulbenchmarkagainstwhichanyalgorithmmaybecompared. X=1[P(min(~c)jf;m;~a) P(min(~c)+1jf;m;~a)]: jyj NoteinparticularthatformanycostfunctionscostvaluesaredistributedGaussianly.For 19
20 suchacase,ifthemeanandvarianceofthegaussianareandrespectively,then()= whichresultina~cwhoseminimumexceedsisgivenby ministic)algorithma,p(~c>jf;m;a)iseither1or0.thereforethefractionofalgorithms erfc(( )=p2)=2,whereerfcisthecomplimentaryerrorfunction. Tocalculatethethirdperformancemeasure,notethatforxedfandm,forany(deter- PaP(min(~c)>jf;m;a) j~c)pap(~cjf;m;a).howevertheratioofthisquantitytopa1isexactlywhatwe Expandingintermsof~c,wecanrewritethenumeratorofthisratioasP~cP(min(~c)> Pa1 : (15)).Thisestablishesthefollowing: Theorem:Forxedfandm,thefractionofalgorithmswhichresultina~cwhoseminimum calculatedwhenweevaluatedmeasureii)(seethebeginningoftheargumentderivingeq. exceedsisgivenbythequantityontheright-handsidesofeqs.(15)and(16). than1/2.forsuchascenario,youralgorithmhasdoneworsethanoverhalfofallsearch ofthe~cproducedinaparticularrunofyouralgorithm,thequantitygivenineq.(16)isless algorithms,forthefandmathand. Soinparticular,considerthescenariowhere,whenevaluatedforequaltotheminimum wellthealgorithm'sperformancecomparestothatoftherandomalgorithm. asmincreases.hereweareinterestedinwhether,asmgrows,thereisanychangeinhow Saythepopulationgeneratedbythealgorithmaaftermstepsisd,anddeney0 Finally,wepresentameasureexplicitlydesignedto\track"analgorithm'sperformance valueofthisnumberofstepsis1 searchalgorithmtosearchx dxandndapointwhoseywaslessthany0.theexpected thatf(x)<y0.nowwecanestimatethenumberofstepsitwouldhavetakentherandom min(~c(d)).letkbethenumberofadditionalstepsittakesthealgorithmtondanxsuch algorithm,onaverage. f(x)<y0.thereforek+1 1=z(d)ishowmuchworseadidthanwouldhavetherandom Sonowimaginelettingarunformanystepsoversometnessfunctionf.Wewishto z(d) 1,wherez(d)isthefractionofX dxforwhich increased.considerthestepwhereandsitsn'thnewvalueofmin(~c).forthatstep, indicatethatsteponourplotasthepoint(n;k+1 1=z(d)).Putdownasmanypointson thereisanassociatedk(thenumberofstepsuntilthenextmin(~c))andz(d).accordingly, makeaplotofhowwelladidincomparisontotherandomalgorithmonthatrun,asm algorithm,thenallthepointsintheplotwillhavetheirordinatevaluesliebelow0.ifthe randomalgorithmwonforanyofthecomparisonsthough,thatwouldmeanapointlying ourplotastherearesuccessivevaluesofmin(~c(d))intherunofaoverf. above0.ingeneral,evenifthepointsalllietoonesideof0,onewouldexpectthatas Ifthroughouttherunaisalwaysabetter\match"tofthanistherandomsearch thesearchprogressesthereiscorresponding(perhapssystematic)variationinhowfaraway 20
21 from0thepointslie.thatvariationtellsonewhenthealgorithmisenteringharderoreasier partsofthesearch. generatemanyoftheseplotsandthensuperimposethem.thiswouldallowyoutoplotthe onecouldreplacethesinglenumberz(d)characterizingtherandomalgorithmwithafull meanvalueofk+1 1=z(d)asafunctionofnalongwithanassociatederrorbar.(Similarly, Notethatevenforaxedf,byusingdierentstartingpointsforthealgorithmonecould 7distributionoverthenumberofrequiredstepstondanewminimum.) functions.thetime-dependentfunctionsweareconcernedwithstartwithaninitialcost Hereweestablishasetofnofreelunchresultsforacertainclassoftime-dependentcost Time-dependentcostfunctions functionthatispresentwhenwesampletherstxvalue.thenjustbeforethebeginning abijectionbetweenfandf.(notethemappinginducedbytfromftofcanvarywith ofeachsubsequentiterationofthesearchalgorithm,thecostfunctionisdeformedtoanew duringthesamplingoftheithpointasfi+1=ti(fi).weassumethatateachstepi,tiis function,asspeciedbythemappingt:fn!f.3wewritethefunctionpresent theiterationnumber.)ifthisweren'tthecase,theevolutionofcostfunctionscouldnarrow inonaregionoff'sforwhichsomealgorithm,\byluck"asitwere,happenstosamplex twodierentpopulationsofyvalues.asbefore,thepopulationdymisanorderedsetofy ityofthesearchalgorithm.ingeneraltherearetwohistogram-basedschemes,involving valuesthatlieneartheextremizingx. valuescorrespondingtothexvaluesindxm.theparticularyvalueindymmatchingaparticularxvalueindxmisgivenbythecostfunctionthatwaspresentwhenxwassampled. Onedicultywithanalyzingtime-dependentcostfunctionsishowtoassessthequal- ff1(dxm(1));;tm 1(fm 1)(dxm(m))g.Similarly,wehaveDym=fTm 1(fm 1)(dxm(1));;Tm 1(fm 1)(dxm(m foreachofthexvaluesindxm.formallyifdxm=fdxm(1);;dxm(m)gthenwehavedym= Incontrast,thepopulationDymisdenedtobetheyvaluesfromthepresentcostfunction thetimescaleoftheevolutionofthecostfunction.insuchsituationsitmaybeappropriate previouselementsofthepopulationarestillalive,andthereforetheir(current)tnessisof tojudgethequalityofthesearchalgorithmwiththehistograminducedbydym;allthose Insomesituationsitmaybethatthemembersofthepopulation\live"foralongtime,on timescaleofevolutionofthecostfunction,onemayinsteadbeconcernedwiththingslike kindofsituation,itmaymakemoresensetojudgethequalityofthesearchalgorithmwith howwellthelivingmember(s)ofthepopulationtrackthechangingcostfunction.inthat interest.ontheotherhand,ifmembersofthepopulationliveforonlyashorttimeonthe thehistograminducedbydym. toaverageoverallpossiblewaysacostfunctionmaybetime-dependent,i.e.,wewishto avengeoverallt(ratherthanoverallf,asinthenfltheorem).soconsiderthesum 3AnobviousrestrictionwouldbetorequirethatTdoesn'tvarywithtime,sothatitisamappingsimply HerewederiveNFLresultsforbothcriteria.InanalogywiththeNFLtheorem,wewish fromftof.ananalysisfort'slimitedthiswayisbeyondthescopeofthispaperhowever. 21
22 astherstmemberofthepopulationisconcerned.soconsideronlyhistogramsconstructed inform>1,andsincef1isxed,thereareaprioridistinctionsbetweenalgorithmsasfar PTP(~cj;f1;T;m;a)wheref1istheinitialcostfunction.NoterstthatsinceTonlykicks fromthoseelementsofthepopulationbeyondtherst.wewillprovethefollowing: Theorem:Forall~c,m>1,algorithmsa1anda2,andinitialcostfunctionsf1, XTP(~cjf1;T;m;a1)=XTP(~cjf1;T;m;a2): Wewillshowthatthisresultsholdswhether~cisconstructedfromdymorfromDym.InanalogywiththeproofoftheNFLtheorem,wewilldothisbyestablishingthea-independence WewillbeginbyreplacingeachTinthesumwithasetofcostfunctions,fi,onefor XTP(~cjf;T;m;a)=XTXdxmX (17) ofptp(~cjf;t;m;a). eachiterationofthealgorithm.todothis,westartwiththefollowing: =XdxmX f2fmp(~cj~f;dxm)p(dxmj~f;m;a) P(f2fm;dxmjf1;T;m;a) f2fmp(~cj~f;dxm;t;m;a) wherewehaveindicatedthesequenceofcostfunctions,fi,bythevector~f=(f1;;fm). XTP(f2fmjf1;T;m;a); formally,usingfi+1=ti(fi),wewrite theseriesisoverthevaluestcantakeforoneparticulariterationofthealgorithm.more NextwedecomposethesumoverallpossibleTintoaseriesofsums.Eachsumin XTP(~cjf;T;m;a)=XdxmX XT1(f2;T1(f1))X f2fmp(~cj~f;dxm)p(dxmj~f;m;a) (NotethatPTP(~cjf;T;m;a)isindependentofthevaluesofTi>m 1,sowecanabsorbthose Tm 1(fm;Tm 1(Tm 2(T1(f1)))): valuesintoanoveralla-independentproportionalityconstant.) numberofbijectionsoffthatmapthatxedcostfunctiontofm.thisisjustaconstant, indicest1:::tm 2.NowforxedvaluesoftheoutersumindicesTm 1(Tm 2(T1(f1))) isjustsomexedcostfunction.accordinglytheinnermostsumovertm 1issimplythe Nowlookattheinnermostsum,overTm 1,forsomexedvaluesoftheoutersum (jfj 1)!. 22
23 SowecandotheTm 1sum,andarriveat XTP(~cjf;T;m;a1)/XdxmX XT1(f2;T1(f1))X f2fmp(~cj~f;dxm)p(dxmj~f;m;a) Tm 1.Infact,allthesumsoverallTicanbedone,leavinguswith NowwecandothesumoverTm 2,intheexactsamemannerwejustdidthesumover Tm 2(fm 1;Tm 2(Tm 3(T1(f1)))): XTP(~cjf;T;m;a1)/XdxmX =XdxmXf2fmP(~cj~f;dxm)P(dxmj~f;m;a) (Inthelaststepwehaveexploitedthestatisticalindependenceofdxmandfm.) ToproceedfurtherwemustdecideifweareinterestedinhistogramsformedfromDymor f2fmp(~cj~f;dxm)p(dxmjf1fm 1;m;a): (18) dym.webeginwithanalysisofthedymcase.forthiscasep(~cj~f;dxm)=p(~cjfm;dxm),since Dymonlyreectscostvaluesfromthelastcostfunction,fm.Pluggingthisinweget XTP(~cjf;T;m;a1)/XdxmX histogramcfromcostvaluesdrawnfromfm.thisconstantwillinvolvethemultinomial Thenalsumoverfmisaconstantequaltothenumberofwaysofgeneratingthe f2fm 1P(dxmjf1fm 1;m;a)XfmP(~cjfm;dxm) theadependence. coecientm theparticulardxm.becauseofthiswecanevaluatethesumoverdxmandtherebyeliminate c1cmandsomeotherfactors.theimportantpointisthatitisindependentof ThiscompletestheproofofEq.(17)forthecasewhere~cisconstructedfromDym. XTP(~cjf;T;m;a)/ f2fm 1XdxmP(dxmjf1fm 1;m;a)/1 X considerablymoredicultsincewecannotsimplifyp(~cj~f;dxm)andthuscannotdecouple thesumsoverfi.nevertheless,thenflresultstillholds.toseethiswebeginbyexpanding Eq.(18)overpossibledymvalues. NextweturnthecasewhereweareinterestednotinDymbutindym.Thiscaseis XTP(~cjf;T;m;a)/XdxmX =XdymP(~cjdym)XdxmX f2fmxdymp(~cjdym)p(dymj~f;dxm) P(dxmjf1fm 1;m;a) myi=1(dym(i);fi(dxm(i))) f2fmp(dxmjf1fm 1;m;a) 23 (19)
24 areleftwithxtp(~cjf;t;m;a)/xdymp(~cjdym)xdxmx term.soitcontributespfm(dym(m);fm(dxm(m))).thisisaconstant,equaltojyjjxj 1.We Thesumovertheinner-mostcostfunction,fm,onlyhasaneectonthe(dym(i);fi(dxm(i))) m 1 Yi=1(dym(i);fi(dxm(i))): f2fm 1P(dxmjf1fm 1;m;a) Thesumoverdxm(m)isnowtrivial,sowehave XTP(~cjf;T;m;a) /XdymP(~cjdym)X m 1 Yi=1(dym(i);fi(dxm(i))): dxm(1)x dxm(m 1)X f2fm 1P(dxm 1jf1fm 2;m;a) mannertotheschemeweusedtoevaluatethesumsoverfmanddxm(m)thatexistedin remainingpopulationofsizem 1ratherthanm.Consequently,inanexactlyanalogous Eq.(19),wecanevaluateoursumsoverfm 1anddxm(m 1).Doingsosimplygenerates NownotethattheaboveequationisoftheexactsameformasEq.(19),onlywitha morea-independentproportionalityconstants.continuinginthismanner,weevaluateall thesumsoverthefiandarriveat Nowthereisstillalgorithm-dependenceinthisresult.Howeveritisatrivialdependence; XTP(~cjf;T;m;a1)/XdymP(~cjdym)X dxm(1)p(dxm(1)jm;a)(dym(1);f1(dxm(1))): rithms.(alternatively,wecouldconsiderallpointsinthepopulation,eventherst,and aspreviouslydiscussed,itarisescompletelyfromhowthealgorithmselectstherstxpoint stillgetannflresult,ifinadditiontosummingoveralltwesumoverallf1.)soeven initspopulation,dxm(1).sinceweconsideronlythosepointsinthepopulationthatare inthecasewhereweareinterestedindymthenflresultstillshold,subjecttotheminor generatedsubsequenttotherst,ourresultsaysthatthereisnodistinctionsbetweenalgo- basedondymordym.forexample,onemaywishtonotconsiderhistogramsatall;onemay caveatsdelineatedabove. judgethequalityofthesearchbythetnessofthemostrecentmemberofthepopulation. Thereareotherswayofassessingthequalityofthesearchalgorithmbesideshistograms algorithmsasfarasthisquantityisconcerned. determinepfp(~cjf;t;m;a).infact,ingeneraltherecanbeaprioridistinctionsbetween onemaywishtocharacterizewhattheaspectsareoftherelationshipbetweenaandtthat Similarly,thereareothersumsonecouldlookatbesidesthoseoverT.Forexample, 24
25 Ximplicitlytakentobeacontiguoussetofintegers).ForthisT,ifaisthealgorithmthat rstsamplesfatx1,nextatx1+1,etc.,regardlessofthevaluesinthepopulation,thenfor theshiftoperator,replacingf(x)byf(x 1)forallx(withmin(x) 1max(x),andwith Asanexampleofsuchdistinctions,saythatforalliterationsofthesearchalgorithm,Tis ~c.sopfp(~cjf;t;m;a)isnotindependentofaingeneral. searchalgorithms,evenforthesameshiftt,thereisnotthisrestrictiononthesetofallowed PfP(~cjf;T;m;a)=0forany~ccontainingcountsinmorethanoneYvaluebin.Forother anyf,thehistograminducedbydymisalwaysmadeupofidenticalyvalues.accordingly, samplesatx1+1,exactlylikealgorithma.ontheotherhand,ifthatvalueishigh,it algorithmlooksattheyvalueoftheitsrstsamplepointx1,andifthatvalueislow,it samplessomepointotherthanx1+1.ingeneral,ifone'sgoalistondminimalyvalues, Indeed,considerthesameshiftT,butusedwithadierentalgorithm,^a.Thisnew 8^acanbeexpectedtooutperforma,evenwhenoneaveragesoverallf. OneobviousdicultywiththeNFLresultsdiscussedaboveisthatonecanalwaysargue\oh, wellintherealworldp(f)isnotuniform,sothenflresultsdonotapply,andtherefore Fixedcostfunctionresults I'mokayinusingmyfavoritesearchalgorithm".Ofcourse,thepremisedoesnotfollowfrom notjustifyanalgorithm.inessence,youmustinsteadmakethemuchbiggerassumption poorlysuitedasoneforwhichitiswellsuited.simplyassumingp(f)isnotuniformcan theproposition.uniformp(f)isatypicalp(f).(theuniformaverageofallp(f)isthe thatp(f)doesn'tfallintothehalfofthespaceofallp(f)inwhichyouralgorithmperforms uniformp(f).)sotheactualp(f)mightjustaseasilybeoneforwhichyouralgorithmis (!)legitimatewayofdefendingaparticularsearchalgorithmagainsttheimplicationsofthe worsethantheuniformp(f). NFLtheorems. ularp(f),andthenarguethatyouralgorithmiswellsuitedtothatp(f).thisistheonly Ultimately,theonlywaytojustifyone'ssearchalgorithmistoargueinfavorofapartic- sweepingthanthenflresults,theseresultsholdnomatterwhattherealworld'sdistribution averagingoverthosesearchalgorithmswhilekeepingthecostfunctionxed.althoughless P(f).Certainsuchresultsapplytowaysofchoosingbetweensearchalgorithms,andinvolve Nonetheless,itisclearlyofinteresttoderiveNFL-typeresultsthatareindependentof examinestwopopulationsdandd0,producedbyaanda0respectively,andbasedonthose overcostfunctionsis. populations,decidestouseeitheraora0forthesubsequentpartofthesearch.asanexample, onechoosingprocedureistochooseaifandonlytheleastcostelementindhaslowercost Letaanda0betwosearchalgorithms.Denea\choosingprocedure"asonethat thantheleastcostelementind0.asanotherexample,a\stupid"choosingprocedurewould chooseaifandonlytheleastcostelementindhashighercostthantheleastcostelement Atthepointthatyouuseachoosingprocedure,youwillhavesampledthecostfunction 25
26 reasons,wecanassumethatthesearchalgorithmchosenbythechoosingproceduredoes thehistogramc>mwhichisthehistogramformedfromd>m.inaddition,foralltheusual thatcomeafterusingthechoosingalgorithm,thenthehistogramtheuserisinterestedinis atallthepointsind[d[d0.accordingly,ifd>mreferstothesamplesofthecostfunction function,observinghowwellanalgorithmhasdonesofartellsusnothingabouthowwellit notreturntoanypointsind[,withoutlossofgenerality4. woulddoifwecontinuetouseitonthesamecostfunction.(forsimplicity,weonlyconsider forusinganyparticularchoosingalgorithm.looselyspeaking,nomatterwhatthecost Thefollowingtheorem,proveninappendixC,tellsuswehavenoapriorijustication deterministicalgorithms.) Theorem:Letdandd0betwoxedpopulationsbothofsizem,thataregeneratedwhen dierentchoosingprocedures.letkbethenumberofelementsinc>m.then thealgorithmsaanda0respectivelyarerunonthecostfunction.letaandbbetwo (Itisimplicitinthistheoremthatthesumexcludesthosealgorithmsaanda0thatdonot Xa;a0P(c>mjf;d;d0;k;a;a0;A)=Xa;a0P(c>mjf;d;d0;k;a;a0;B): (20) resultindandd0respectivelywhenrunonf.) equally,whenforanygivenfsomepopulationswillbemorelikelythanothers.howevereven ifoneweightspopulationsaccordingtotheirprobabilityofoccurrence,itisstilltruethat, onaverage,thechoosingprocedureoneuseshasnoeectonlikelyc>m.thisisestablished Onemightthinkthattheprecedingtheoremismisleading,sinceittreatsallpopulations bythefollowingcorollary. Corrolary:Undertheconditionsgivenintheprecedingtheorem, Proof:Let\proc"refertoourchoosingprocedure.Weareinterestedin Xa;a0P(c>mjf;m;k;a;a0;A)=Xa;a0P(c>mj;f;m;k;a;a0;B): (21) Xa;a0P(c>mjf;m;k;a;a0;proc)= a;a0;d;d0p(c>mjf;d;d0;k;a;a0;proc) X ithasn'tseenyetbutthata0has(andvice-versa).ratherthanhavethedenitionofasomehowdepend ontheelementsind0 d(andsimilarlyfora0),wedealwiththisproblembydeningc>mtobesetonlyby 4acanknowtoavoidtheelementsithasseenbefore.Howeverapriori,ahasnowaytoavoidtheelements P(d;d0jf;k;m;a;a0;proc): populationd>m. thoseelementsind>mthatlieoutsideofd[.(thisissimilartotheprocedurewedevelopedabovetodeal d[aswellasofd>m.italsomeanstheremaybefewerelementsinthehistogramc>mthanthereareinthe withpotentiallyretracingalgorithms.)formally,thismeansthattherandomvariablec>misafunctionof 26
27 (i.e.,anyparticularpairofvaluesofdandd0).forthatterm,p(d;d0jf;k;m;a;a0;proc) Pullthesumoverdandd0outsidethesumoveraanda0.Consideranyterminthatsum otherwise.(recallthatweareassumingthataanda0aredeterministic.)thismeansthat isjust1forthoseaanda0thatresultindandd0respectivelywhenrunonf,and0 overdandd0isthesameforchoosingproceduresaandb.thereforethefullsumisthe sameforbothprocedures.qed. consideredinourtheorem.accordingly,ourtheoremtellusthatthesummandofthesum thep(d;d0jf;k;m;a;a0;proc)factorsimplyrestrictsoursumoveraanda0totheaanda0 onewillbechoosingamong. choosingprocedure,onemusttakeintoaccountnotonlyp(f)butalsothesearchalgorithms somechoosingprocedureasfarassubsequentsearchisconcerned.tohaveanintelligent TheseresultstellusthatthereisnoassumptionforP(f)that,byitself,justiesusing thatforxedf1andf2,iff1doesbetter(onaverage)withthealgorithmsinsomeset A,thenf2doesbetter(onaverage)withthealgorithmsinthesetofallotheralgorithms. proceduresafalwaysusealgorithmag,andbfalwaysusealgorithma0g.thiscasemeans Theseresultsalsohaveinterestingimplicationsifoneconsidersthe\degenerate"choosing performancethandoestherandomf,thenthatwell-behavedfgivesworsethanrandom Inparticular,ifforsomefavoritealgorithmsacertain\well-behaved"fresultsinbetter behavioronthesetallremainingalgorithms. P(f),thenstupidchoosingprocedures{likechoosingthealgorithmwiththelessdesirable~c relatedtothetheoremabove[16].translatedintothecurrentcontextthatresultsuggests thatifonerestrictsthesumstoonlybeoverthosealgorithmsthatareagoodmatchto Infact,thingsmayverywellbeworsethanthis.Insupervisedlearning,thereisaresult tobesuperiortoadumboneisbeyondthescopeofthispaper.butclearlytherearemany ofwhatexactlythesetofalgorithmssummedovermustbeforasmartchoosingprocedure subtleissuestodisentangle. {outperform\smart"ones(whicharetheoneseveryoneusesinpractice).aninvestigation DiscussionandFutureWork Inthispaperwepresentaframeworkforinvestigatingsearch.Thisframeworkservesasa \skeleton"forthesearchproblem;ittellsuswhatwecanknowaboutsearchbefore\eshing in"thedetailsofaparticularrealworldsearchproblem.phraseddierently,itprovidesa aboutthem. languageinwhichtodescribesearchalgorithms,andinwhichtoask(andanswer)questions specicallytailoredtomatchthosefeatures.theinverseprocedure farmorepopular givenf,determinecertainsalientfeaturesofit,andthenconstructasearchalgorithm,a, formygivencostfunctionf?"theproperanswertothisquestionistostartwiththe Ultimately,ofcourse,theonlyimportantquestionis,\HowdoIndgoodsolutions insomecommunities istoinvestigatehowspecicalgorithmsperformondierentf's. 27
28 P(f).Tounderstandthis,rstnotethatwedoinfactknowfexactly.Butatthesame procedure,ofgoingfrom(featuresconcerning)ftoanappropriatea. Thisinverseprocedureisonlyofinteresttothedegreethatithelpsuswithourprimary time,thereismuchaboutfthatweneedtoknowthatiseectivelyunknowntous(e.g., f'sextrema).inthis,itisasthoughfispartiallyunknown.theverynatureofthesearch Notethatoftenthe\salientfeatures"concerningfcanbestatedintermsofadistribution paper. ndingagoodaforaparticularp(f)-exactlytheissueaddressedinsection3ofthis processistoadmitthatyoudon't\know"finfull.asaresult,itmakessenseto(implicitly orotherwise)replacefwithadistributionp(f).inthis,thesearchproblemreducesto andgeneticalgorithms)areunabletocompetewithcarefullyhand-craftedsolutionsfor specicsearchproblems.thetravelingsalesmanproblem(tsp)isanexcellentexample ofsuchasituation;thebestsearchalgorithmsforthetspproblemarehand-tailoredforit Asanexampleofallthis,itiswellknownthatgenericmethods(likesimulatedannealing [12].Linearprogrammingproblemsareanotherexample;thesimplexalgorithmisasearch concerningfandtherebyeectivelyreplacefwithap(f);andthenuseasearchalgorithm situations,theprocedurefollowedbytheresearcheristo:identifysalientaspectsoff(e.g., itisatspproblem,oritisalinearprogrammingproblem);throwawayallotherknowledge algorithmspecicallydesignedtosolvecostfunctionsofaparticulartype.inbothofthese explicitlyknowntoworkwellforthatp(f). pretendthatonesimplyhasageneraltspproblem particularsunknown andusean hasaparticulartravelingsalesmanproblem(tsp)problemathand,onewouldinstead itsextremaaren'tknown),andthereforeonereplacesitwithap(f).forexample,ifone Inotherwords,oneadmitsthatinacertainsensefisnotcompletelyknown(forexample, questionweaddressedwaswhetheritmaybethatsomealgorithmaperformsbetterthan B,onaverage.Ouranswertothisquestion,givenbytheNFLtheoremisthatthisis algorithmwell-suitedtotspproblemsingeneral. impossible.animportantimplicationofthisresultisthe\conservation"natureofsearch, Inourinvestigationofthesearchproblemfromthismatch-a-to-fperspective,therst illustratedbythefollowingexample.ifageneticalgorithmoutperformssimulatedannealing oversomeclassofcostfunctions,thenovertheremainingcostfunctionsfn,simulated annealingmustoutperformthegeneticalgorithm.itshouldbenotedthatthisconservation appliesevenifoneconsiders\adaptive"searchalgorithms[6,18]whichmodifytheirsearch featuresoff. strategybasedonpropertiesofthepopulationof(x Y)pairsobservedsofarinthesearch, andwhichperformthis\adaptation"withoutregardtoanyknowledgeconcerningsalient isviewedasoptimizationoveracostor\tness"function.wefurthersimplifymattersby algorithms).tothisend,considertheextremelysimpliedviewinwhichnaturalselection relationshipbetweennaturalselectioninthebiologicalworldandoptimization(i.e.genetic Itisimportanttobearinmindexactlywhatallofthisdoes(not)implyaboutthe assumingthetnessfunctionisstaticovertime. sinceitbegan,andthereforewedon'tallowanalgorithmtoresamplepointsithadalready Inthispaperwemeasureanalgorithm'sperformancebasedonallXvaluesithassampled 28
29 evolutionthroughtimeof\generations"consistingoftemporallycontiguoussubsetsofour onemightconsiderdierentmeasures.inparticular,wemaybeprimarilyinterestedinthe population,generationsthatareupdatedbyoursearchalgorithm. visited.ournfltheoremstatesthatallalgorithmsareequivalentbythismeasure.however NFLtheoremdoesnotapplytothisalternativekindofperformancemeasure.Forexample, accordingtothisalternativeperformancemeasure,analgorithmthatresamplesoldpoints inxthataretandaddsthemtothecurrentgenerationwillalwaysdobetterthanone Insuchascenario,itdoesmakesensetoresamplepointsalreadyvisited.Moreover,our selectionmeansthatonly(essentialcharacteristicsof)goodpointsinxarekeptaroundfrom kindofmeasure;weonlyseetheorganismsfromthecurrentgeneration.inaddition,natural thatresamplesoldpointsthatarenott. onegenerationtothenext.accordingly,usingthissecondkindofperformancemeasure, Nowwhenweexaminethebiologicalworldaroundus,weareimplicitlyusingthissecond theenvironment-i.e.,costfunction-didn'tchangeintime,etc.)thisisnothingmorethan thetautologythatnaturalselectionimprovesthetnessofthemembersofageneration. oneexpectsthattheaveragetnessacrossagenerationimproveswithtime.(orwouldif notmeanthatifwewishtodoasearch,andareabletokeeparoundallpointssampledso performswellaccordingtothisgeneration-basedmeasuredoesnotmeananythingconcerning itsperformanceaccordingtothe~c-basedmeasureusedinthispaper.inparticular,itdoes Howevertheevidencegarneredfromexaminingtheworldaroundusthatnaturalselection Yetitispreciselythissituationthatisofinterestintheengineeringworld. far,thatwehaveanyreasontobelievethatnaturalselectionisaneectivesearchstrategy. selectionisaneectivesearchstrategyinthebiologicalworld.wesimplyhavenothada chancetoobservethebehaviorofalternativestrategies.accordingtothenfltheorem,for thatnaturalselectionisaneectivesearchstrategy.itdoesnotevenindicatethatnatural Inshort,theempiricalevidenceofthebiologicalworlddoesnotindicateinanysense (Thisisexactlyanalogoustothefactthathill-descendingcanbeathill-climbingatnding allweknow,thestrategyofbreedingonlytheleasttmembersofthepopulationmayhave tnessmaxima.)thebreed-the-worststrategywillingeneralresultinworserecentgenerations,butsimplythefactthatyouareusingthatstrategyimpliesnothingaboutthequality ofthepopulationsoverthelongterm. selection,onewouldhavetoallowthebreed-the-worststrategytoexploitthesamemassive amountofparallelismexploitedbynaturalselectionintherealworld,wheretherearea hugenumberofgenomesevolvinginparallel.itmaywellbethatthe\blindwatchmaker" Inthisregard,notethattofairlycomparethebreed-the-worststrategywithnatural doneabetterjobatndingtheextremaofthecostfunctionfacedbybiologicalorganisms. hasmanagedtoproducesuchanamazingbiomesimplybyrelyingonmassiveparallelism ratherthanbreed-the-best.nobodyknows;nobodyhastriedtomeasure\howwell"natural butbreed-the-worstwinsinothers. themeasurementsarenallydonewewillndthatnaturalselectionwinsinsomeecosystems selectionvs.breed-the-worstvariesfromecosystemtoecosystem itmaywellbethatwhen selectionworksinthebiologicalworldbefore.indeed,presumablytheecacyofnatural Ontheotherhand,ifwerelaxtheunrealisticassumptionthatthetnessfunctioniscon- 29
30 ratherthanabreed-the-worststrategy,regardlessoftheecosystem.(suchadvantagescould thatthe\matching"ofsearchalgorithmandcostfunctionrequiredbytheinnerproduct arisefromthefactthatthecostfunctionisbeingdeterminedinpartbythepopulation,so stantovertime,thenitispossiblethattheremaybeadvantagestousingnaturalselection outthatbreed-the-worsthasadvantagesovernaturalselectionforvaryingtnessfunctions vantagesrelativetonaturalselection'sbreed-the-beststrategy.alternatively,itmayturn and/orminimaxconcerns.theseareissuesforfutureresearch. formulamaysomehowbeautomatic.)similarly,thatstrategymayhaveminimaxdisad- betweenthetwosearchalgorithms.thisraisessomeobviousquestionsforfutureresearch: worstmembersofthepopulationforthenextgenerationisequivalenttoonethatkeepsthe bestmembers,onaverage.however,thetnessofthemembersofthegenerationswilldier Tosummarize,bytheNFLtheorem,anygeneration-basedschemethatkeepsonlythe Averagedoverallf,howbigwouldoneexpectthedierencetobe?Foraxedf,andtwo thepopulationwill(likely)beforarandomalgorithmasmgrows? thislastcalculationcomparewiththecalculationmadeaboveofwhatthebestmemberof beinginthecurrentgeneration,howbigwouldoneexpectthedierencetobe?howdoes identicalrandomsearchalgorithmsthatare\directed"dierentlyinwhotheyclassifyas Itisperhapsttingforapaperabouteectivesearchthatweconcludewithabrieflisting 9.2 ofother(research)directionswebelievewarrantfurtherinvestigation. Futurework tooltosolverealproblems.thiswouldinvolvetwosteps.firstweneedamethodof haveusedp(f)todothis,butperhapsthereareotherwaysthatweshouldalsoconsider. incorporatingbroadkindsofknowledgeconcerningfintotheanalysis.inthispaperwe Themostimportantcontinuationofthisworkistoturnourframeworkintoapractical theknowledgeconcerningthecostfunctionthatisimplicitintheheuristicsofbranchand throughtheassemblageofsub-solutions? Boundstrategies.Howdoweincorporatehowthecostofacompletesolution(f)isaccrued Forexample,itisnotyetclearhowto(orevenwhetheroneshould)encapsulateinaP(f) concerningfintoanoptimala.thegoalinitsbroadestsenseistodesignasystemthatcan takeinsuchknowledgeconcerningfandthensolvefortheoptimalagiventhatknowledge. (Forexample,iftheknowledgewereintheformofP(f),onewould\invert"theinner Thesecondstepinthissuggestedprogramistodeterminehowbesttoconvertknowledge onlythetoolsdevelopedinthispaper.manyofthemwerepresentedinthetext.others, therearemanyimportantquestionsrelatedtothisprogramthatshouldbeanalyzableusing productformulasomehow.)onewouldthenusethatatosearchthef. particularlywell-suitedtohelpusunderstandtheconnectionbetweenp(f)andanoptimal Initsfullestsense,thisprogrammaywellinvolvemanyyearsofwork.Nonetheless, thediagonalinfspace(i.e.,frombeinguniformoverallf),howwillcertaina'sbehurt a,are:howfastdoesthecosthistogram~cassociatedwithaparticularalgorithmconverge tothehistogramofthecostvaluesftakesonacrossallofx?asp(f)changesfrom andcertaina'shelped?couldtheaverageoveralla'simprove?forwhatp(f)'sbesides 30
31 algorithms),forwhatp(f)istheperformanceofthealgorithmsequal?inparticular,if thediagonalareallalgorithmsequal?giventwoparticularalgorithms(ratherthanall P(f)isuniformoversomesubsetFandzerooutside,5whataretheequivalence classesofsearchalgorithmswithidenticalexpectedbehavior? populationcanonlyimprove.soallpreviousstudiesshowingthattnessdoesimprove above.foranyalgorithm,asthesearchprogresses,thetnessofthebestmemberofthe currentlypopularsearchalgorithmsintermsoftheperformancebenchmarkswepresent Asapreliminarystepinthisprogram,itwouldmakesensetoexploretheecacyof bettertheimprovementisthanyouwouldexpectittobesolelyduetothe\ttestcanonly improve"eect.that'swhatourmeasuresaredesignedtoassess. intimeforsomealgorithmareallydon'tproveanything.what'simportantishowmuch rangeofpopulationsizes.thingsshouldbeevenworseifonerandomlysamplesfromthe quitelikelythatonasignicantfractionoftheproblemsinthestandardtestsuites,oneor moreofthecurrentlypopularsearchalgorithmswillfailtoperformwell,atleastforsome Giventherecentexperienceinthesupervisedlearningcommunity[8,13,10],itseems spaceofreal-worldsearchproblems.thisisbecausethereare\selectioneects"ensuring thatthemostcommonlystudiedsearchproblems(i.e.,thoseinthesuites)arethosewhich peopleconsider\reasonable";inpractice,\reasonable"oftensimplymeans\agoodmatch ministicalgorithms.aretherepotentialadvantagestostochasticalgorithms?inparticular, tothealgorithmsi'mfamiliarwith". gorithmsa?i.e.,canonewritep(cjf;m;)=paka;p(cjf;m;a)forsomeexpansion doesitmakesenseto\expand"anystochasticalgorithmintermsofdeterministical- Anotherinterestingseriesofquestionsconcernsdierencesbetweenstochasticanddeter- coecientska;?ifso,itsuggeststhatasp(f)movesfromthediagonaltheperformance algorithmshavecertainminimaxadvantagesoverdeterministicones. of'swillneitherimprovenordegradeasmuchasthatofa's.soitmaybethatstochastic distinctionsoccurin\cycles",inwhichalgorithmais(head-to-headminimax)superior tob,andbtoc,butthencisalsosuperiortoa.argumentsforchoosingbetween minimaxdistinctionsbetweenalgorithms.perhapsthesimplestistocharacterizewhensuch Therearemanyotherissuesthatremaintobeinvestigatedconcerninghead-to-head algorithmsbasedonhead-to-headminimaxdistinctionsaremorepersuasiveintheabsence withtheexample)forsomereasonalgorithmccanberuledoutasacandidatealgorithm ofsuchcycles.howeveritshouldbenotedthateveniftherearesuchcycles,if(tocarryon (e.g.,ittakestoolongtocompute,orisdiculttodealwith,orsimplyisnotinvogue),then adoptedinthispaperandconventionalstatistics.inparticulartheeldofoptimalexperimentaldesign[1]andmorepreciselyactivelearning[2]isconcernedwiththefollowing Otherissuestobeexploredinvolvetherelationbetweenthestatisticalviewofsearch minimaxdistinctions. thefactthatwehaveacycledoesnotprecludechoosingalgorithmabasedonhead-to-head question:thereissomeunknownprobabilisticrelationshipbetweenxandy.ihaveasetof pairsofx-yvaluesformedbysamplingthatrelationship(the\trainingset").atwhatnext 5Asanexample,mightbethesetofcorrelatedcostfunctionsasin[14]. 31
32 Thisquestionofhowbesttoconductactivelearningisobviouslyverycloselyrelatedtothe searchproblem;futureworkinvolvesseeingwhatresultsintheeldofactivelearningcan XvalueshouldIsampletherelationshipto\best"helpmeinferthefullX-Yrelationship? befruitfullyappliedtosearch. algorithmsaretooneanother.asanexampleofsuchameasure,onecouldsimplysaythat algorithm.accordingly,thisequationprovidesseveralwaystomeasurehow\close"two whatwewanttoknowissetbyit).thersttermontheright-handsideissetbyone's ConsideragainEq.(4).Theleft-handsideiswhatweareinterestedin(ormoregenerally, the(~c-indexed)vectorsp(~cjm;a)areforthosetwoalgorithms,forthatp(f).(onecould onecouldmeasuretheclosenessoftwoalgorithmsforaspecicp(f),byseeinghowclose imaginethatforsomep(f)twoalgorithmswillbeclose,whileforotherstheywillbefar howclosetwoalgorithmsareisgivenbyhowclosetheirvectors~vc;a;mare.alternatively, apart.)asanalexample,givenanalgorithm,onecouldsolveforthep(f)thatoptimizes simulatedannealing,eventhoughitsinternalworkingsarecompletelydierent".onecould fortwoalgorithms,andusethistomeasuretheclosenessofthealgorithmsthemselves. P(~cjm;a)insomenon-trivialsense.OnecouldthenseehowclosetheoptimalP(f)'sare alsoinvestigatehypotheseslike\allalgorithmsthathumansconsider'reasonable'arecloseto oneanother".futureworkinvolvesexploringthesemeasuresoftheclosenessofalgorithms. Withthesekindsofmeasures,onecouldsaythingslike\thisalgorithmisverycloseto changingthesearchalgorithm.thecostfunctiondoesn'tchangewhenwere-encode tothatencoding.howeverinthecontextofthispaper,changingtheencodingmeans duringsearch.normallyonetalksofhowthecostfunctionisencoded,andpossiblechanges Otherfutureworkinvolvesexploringtheimportanceofthe\encoding"schemeoneuses ratherhowwe(thealgorithm)viewthefunctionchanges. encodings"ofcostfunctions.forexample,if(a)isare-encodingofalgorithma,then onemightsaythatacostfunctionfbecomes(f)underthatsamere-encodingip(~cj f;m;a)=p(~cj(f);m;(a))forall~c.(alternatively,onemightsaythat(a)isalegal Nonetheless,onecanimagineseveralwaystocouplere-encodingofalgorithmswith\re- true.)futureworkhereinvolvesseeinghowchangingtheencodingschemeinteractswith P(f)todeterminetheecacyofthesearchprocess. re-encodingschemeforalgorithmsithereisanassociated(f)forwhichtheforegoingis mustbemodied(andhow)ifwestillhavep(f)=xp0(f(x))butnolongerhaveuniform y2y.aninterestingquestionforfutureresearchistoseewhichoftheresultsofthispaper P0(y).(Intuitively,forsuchaP(f),f(x)isbeingsetafteryoupickxasthenextpointto UniformP(f)canberewrittenasP(f)=xP0(f(x)),whereP0(y)isuniformoverall visit,andthisisbeingdonewithoutanyregardforpointsyou'vealreadyseen.hence,one somenearestneighborcoupling? ofalgorithmsareequal?andwhathappensifratherthanequalxp0(f(x)),p(f)involves forwhichallalgorithmsareequal?whatisthemostgeneralp(f)forwhichaparticularpair wouldexpectnfl-resultstohold.)relatedquestionsare:whatisthemostgeneralp(f) thatnotcanbewrittenasxp0(f(x))butforwhichitisstilltruethatallalgorithmsare equal.forexample,sayjyj>jxjandletp(f)bei)uniformoverallfsuchthatforno Inrelationtotherstandlastofthesequestions,itseemsplausiblethatthereareP(f)'s 32
33 innfl-typeresults,sincethepointsyouhaveseensofartellyounothingaboutwhereyou P(f)hasextremelystrongcouplingbetweentheelementsofthepopulation,incontrastto P(f)'sthatcanbewrittenasxP0(f(x)).YetitseemslikelythattheseP(f)'salsoresult x1;x22xdoesf(x1)=f(x2);andii)zeroforallfthatdon'tobeythiscondition.this shouldsearchnext. more\real-world"p(f)andstillhavenfl-typeresults? holdraiseanintriguingquestion:justhowfarcanonepushfromtheuniformp(f)toa practicalconcerns.yetthesebroaderclassesofp(f)'sforwhichnfl-typeresultsmight InthispaperthechoiceofP(f)(uniform)wasmotivatedbytheoereticalratherthan stoppingcondition,afunctionofallpopulationsuptothepresent,ismet.thenintuitively, bynfl,onewouldexpectthataveragedoverallf,theprobabilitythatyouralgorithmstop hadtimetoexplicatehere.forexample,consideralgorithmsthatkeeprunninguntilsome Finally,therearemanyotherNFL-typeresults,foruniformP(f),thatwehavenot (andsimilar)resultsisthesubjectoffuturework. aftermsamplesoffisindependentofthealgorithmbeingused.theformalproofofthese helpfulconversation,andthesfiforfunding.dhwwouldalsoliketothanktxninc.for funding. WewouldliketothankRajaDas,TalGrossman,PaulHelman,andUnamayO'Reillyfor Acknowledgments References [1]J.O.Berger,StatisticalDecisonTheoryandBayesianAnalysis,Springer-Verlag(1985). [3]T.Cover,J.Thomas,ElementsofInformationTheory,JohnWiley&Sons,(1991). [2]D.Cohn,NeuralNetworkExplorationUsingOptimalExperimentalDesign,MITAI Memo [4]M.R.Garey,D.S.Johnson,ComputersandIntractability,Freeman(1979). [6]L.Ingber,AdaptiveSimulatedAnnealing,Softwarepackagedocumentation, [5]J.Holland,AdaptationinNaturalandArticialSystems,UniversityofMichiganPress, AnnArbor,(1975). [7]S.Kirkpatrick,C.D.GelattJr.,M.P.Vecchi,Science,220,671,(1983). ftp.alumni.caltech.edu:/pub/ingber/asa.tar.gz. [8]R.Kohavi,personalcommunication.AlsoseeAStudyofCross-ValidationandBootstrapforAccuracyEstimationandModelSelection,tobepresentedatIJCAI1995. [9]E.L.Lawler,D.E.Wood,OperationsResearch,14(4), ,(1966). 33
34 [11]J.Pearl,Heuristics,intelligentsearchstrategiesforcomputerproblemsolving,Addison- [10]P.Murphy,M.Pazzani,JournalofArticialIntelligenceResearch,1, (1994). [12]GerhardReinelt,TheTravelingSalesman,computationalsolutionsforTSPapplications,SpringerVerlagBerlinHeidelberg(1994). Wesley,(1984). [14]P.F.Stadler,Europhys.Lett.20,pp ,(1992). [13]C.Schaer,ConservationofGeneralization:ACaseStudy. [15]C.E.M.Strauss,D.H.Wolpert,D.R.Wolf.Alpha,Evidence,andtheEntropicPrior [16]DH.Wolpert,O-trainingseterrorandaprioridistinctionsbetweenlearningalgorithms,TechnicalReportSFI-TR ,SantaFeInstitute,1995. (1992). inmaximumentropyandbayesianmethods,ed.alimohammed-djafari,pp , [17]DH.Wolpert,OnOverttingAvoidanceasBias,TechnicalReportSFI-TR , SantaFeInstitute,1992. [18]D.Yuret,M.delaMaza,DynamicHill-Climbing:OvercomingthelimitationsofoptimizationtechniquesinTheSecondTurkishSymposiumonArticialIntelligenceand A NeuralNetworks,pp ,(1993). search Proofrelatedtoinformationtheoreticaspectsof Wewanttocalculatetheproportionofallalgorithmsthatgiveaparticular~cforaparticular butnite-list.thatlistisindexedbyallpossibled's(asidefromthosethatextendover theentireinputspace).eachentryinthelististhextheainquestionoutputsforthat f.weproceedinseveralsteps. d-index. 1)SinceXisnite,populationsarenite.Thereforeany(deterministic)aisahuge- nowonweimplicitlyrestrictthediscussiontounorderedpathsoflengthm.)aparticular is\in"or\from"aparticularfifthereisaunorderedsetofm(x;f(x))pairsidentical thesamexvalue.suchasetisan\unorderedpath".(withoutlossofgenerality,from 2)Consideranyparticularunorderedsetofmx ypairswherenotwoofthepairsshare to.thenumeratorontheright-handsideofeq.(9)isthenumberofunorderedpathsin thegivenfthatgivethedesired~c. ontheright-handsideofeq.(9)-isproportionaltothenumberofa'sthatgivethedesired 3)Claim:Thenumberofunorderedpathsinfthatgivethedesired~c-thenumerator 34
35 ~cforf.(theproofofthisclaimwillconstituteaproofofeq.(9).)furthermore,the ~cforf,andfromitproducesathatisinfandgivesthedesired~c.wewillthenshow proportionalityconstantisindependentoffand~c. of;f,and~c.theproofwillthenbecompletedbyshowingthatissingle-valued,i.e.,by thatforanythenumberofalgorithmsasuchthat(a)=isaconstant,independent 4)Proof:Wewillconstructamapping:a!.takesinanathatgivesthedesired showingthatthereisnoawhohasasimageundermappingmorethanone. Indicatebyd(ord)thissetoftherstmd'sprovidedbyord.(Notethatanyordisitself inturnprovidesasetofmsuccessived's(ifoneincludesthenulld)andafollowingx. (Notethateveryxvalueinanunorderedpathisdistinct.)Eachsuchorderedpathord 5)Anyunorderedpathgivesasetofm!dierentorderedpathsintheobviousmanner. apopulation,buttoavoidconfusionweavoidreferringtoitassuch.) distinctpartiala'sforeach(oneforeachorderedpathcorrespondingto),wehavem! thelistofana,butwithonlythemd(ord)entriesinthelistlledin;theremainingentries areblank.(wesaythatmisthe\length"ofthepartialalgorithm.)sincetherearem! 6)Foranyorderedpathordwecanconstructa\partialalgorithm".Thisconsistsof suchpartiallylled-inlistsforeach. onepartialalgorithmgeneratedfromandthatgive~cwhenrunonf). \consistent"withaparticularfullalgorithm.thisallowsustodene(theinverseof):for anythatisinfandgives~c, 1()(thesetofallathatareconsistentwithatleast 7)Intheobviousmannerwecantalkaboutwhetheraparticularpartialalgorithmis adistinctm-elementpartialalgorithm.ourquestionishowmanyfullalgorithmslistsare rstgenerateallorderedpathsinducedbyandthenassociateeachsuchorderedpathwith give~c, 1()containsthesamenumberofelements,regardlessof,f,orc.Tothatend, 8)Tocompletetherstpartofourproofwemustshowthatforallthatareinfand consistentwithatleastoneofthesepartialalgorithmpartiallists.(howthisquestionis answeredisthecoreofthisappendix.) permutingtheindicesdofallthelists.obviouslysuchareorderingwon'tchangetheanswer toourquestion. 9)Toanswerthisquestion,reordertheentriesineachofthepartialalgorithmlistsby anydindexoftheform((dx(1);dy(1));:::;(dx(im);dy(im)))whoseentryislledin arbitraryconstantyvalueandxjreferstothej'thelementofx.next,createsomearbitrary inanyofourpartialalgorithmlistswithd0(d)((dx(1);z);:::;(dx(i);z)),wherezissome 9)Wewillperformthepermutingbyinterchangingpairsofdindices.First,interchange listswithd00(d0)((x1;z);:::;(xm;z)).(recallthatallthedx(i)mustbedistinct.) butxedorderingofallx2x:(x1;:::;xjxj).theninterchangeanyd0indexoftheform ((dx(1);z;:::;(dx(im);z)whoseentryislledininanyofour(new)partialalgorithm atleastonepartialalgorithmlistin 1()isindependentof,candf.Thiscompletes asisthenumberofsuchlists(it'sm!).thereforethenumberofalgorithmsconsistentwith therstpartoftheproof. 10)Byconstruction,theresultantpartialalgorithmlistsareindependentof,~candf, AandB.ThereisnoorderedpathAordconstructedfromAthatequalsanorderedpath 11)Forthesecondpart,rstchooseany2unorderedpathsthatdierfromoneanother, 35
36 BordconstructedfromB.SochooseanysuchAordandanysuchBord.Iftheydisagreefor them.iftheyagreeforthatd,thentheyhavethesamedouble-elementd.continueinthis theyagreeforthenulld,thensincetheyaresampledfromthesamef,theyhavethesame single-elementd.iftheydisagreeforthatd,thenthereisnoathatagreeswithbothof thenulld,thenweknowthatthereisno(deterministic)athatagreeswithbothofthem.if havedisagreedatsomepointbynow,andthereforethereisnoathatagreeswithbothof them. manneralltheuptothe(m 1)-elementd.Sincethetwoorderedpathsdier,theymust ain 1(A)thatisalsoin 1(B).Thiscompletestheproof. 12)SincethisistrueforanyAordfromAandanyBordfromB,weseethatthereisno B rithms Proofrelatedtominimaxdistinctionsbetweenalgo- Theproofisbyexample. 1)Lettherstpointa1visitsbex1,andtherstpointa2visitsbex2. ConsiderthreepointsinX,x1;x2,andx3,andthreepointsinY,y1;y2,andy3. 3)Ifatitsrstpointa2seesay1,itjumpstox1.Ifitseesay2,itjumpstox3. 2)Ifatitsrstpointa1seesay1oray2,itjumpstox2.Otherwiseitjumpstox3. ConsiderthecostfunctionthathasastheYvaluesforthethreeXvaluesfy1;y2;y3g, respectively. populationcontainingy2andy3andsuchthata2producesapopulationcontainingy1and (y2;y3). Theproofiscompletedifweshowthatthereisnocostfunctionsothata1producesa Form=2,a1willproduceapopulation(y1;y2)forthisfunction,anda2willproduce y2.therearefourpossiblepairsofpopulationstoconsider: ii)[(y2;y3);(y2;y1)]; i)[(y2;y3);(y1;y2)]; iii)[(y3;y2);(y1;y2)]; ay2itssecondpointmustequala2'srstpoint.thisrulesoutpossibilitiesi)andii). Sinceifitsrstpointisay2a1jumpstox2whichiswherea2starts,whena1'srstpointis iv)[(y3;y2);(y2;y1)]. fy3;s;y2g,forsomevariables.forcaseiii),swouldneedtoequaly1,duetotherstpoint Forpossibilitiesiii)andiv),bya1'spopulationweknowthatfmustbeoftheform 36
37 ina2'spopulation.howeverforthatcase,thesecondpointa2seeswouldbethevalueatx1, thereforeseeay2,contrarytohypothesis. whichisy3,contrarytohypothesis. population.howeverthatwouldmeanthata2jumpstox3foritssecondpoint,andwould Accordingly,noneofthefourcasesispossible.Thisisacasebothwherethereisno Forcaseiv),weknowthattheswouldhavetoequaly2,duetotherstpointina2's histograms.qed. symmetryunderexchangeofdy'sbetweena1anda2,andnosymmetryunderexchangeof CSinceany(deterministic)searchalgorithmisamappingfromdDtoxX,anysearch algorithmisavectorinthespacexd.thecomponentsofsuchavectorareindexedbythe ProofrelatedtoNFLresultsforxedcostfunctions possiblepopulations,andthevalueforeachcomponentisthexthatthealgorithmproduces dered)elements.thesetofthosepopulationsthatdostartwithdthiswaydenesasetof otherpopulationofsizegreaterthanmhasthe(ordered)elementsofdasitsrstm(or- giventheassociatedpopulation. componentsofanyalgorithmvectora.thosecomponentswillbeindicatedbyad. Considernowaparticularpopulationdofsizem.Givend,wecansaywhetherany thatareequivalenttotherstm<melementsindforsomem.thevaluesofthose componentsforthevectoralgorithmawillbeindicatedbyad.thesecondtypeconsistsof thosecomponentscorrespondingtoallremainingpopulations.intuitively,thesearepopulationsthatarenotcompatiblewithd.someexamplesofsuchpopulationsarepopulations Theremainingcomponentsofaareoftwotypes.Therstisgivenbythosepopulations indicatedbya?d. thatcontainasoneoftheirrstmelementsanelementnotfoundind,andpopulationsthat re-ordertheelementsfoundind.thevaluesofaforcomponentsofthissecondtypewillbe LetprocbeeitherAorB.Weareinterestedin Xa;a0P(c>mjf;d1;d2;k;a;a0;proc) a?d;a0?d0x =Xad;a0d0X Thesummandisindependentofthevaluesofa?danda0?dforeitherofourtwod's. ad;a0d0p(c>mjf;d;d0;k;a;a0;proc): populationsnotconsistentwithd,ofthenumberofpossiblexeachsuchpopulationcould Inaddition,thenumberofsuchvaluesisaconstant.(Itisgivenbytheproduct,overall bemappedto.)therefore,uptoanoverallconstantindependentofd,d0,f,andproc,our sumequals ad;a0d0x Xad;a0d0P(c>mjf;d;d0;ad;a0d0;ad;a0d0;proc): 37
38 sumreducesto (namely,thevaluethatgivesthenextxelementind),andsimilarlyfora0d0.thereforeour isdened.thismeansthatweactuallyonlyallowonevalueforeachcomponentinad Bydenition,weareimplicitlyrestrictingthesumtothoseaanda0sothatoursummand ad;a0d0p(c>mjf;d;d0;ad;a0d0;proc): X choiceofaora0isxed.accordingly,withoutlossofgenerality,wecanrewriteoursumas isoverthesamecomponentsofaasthesumovera0d0isofa0.nowforxeddandd0,proc's Notethatnocomponentofadliesindx[.Thesameistrueofa0d0.Sooursumoverad withtheimplicitassumptionthatc>missetbyad.thissumisindependentofproc.qed. XadP(c>mjf;d;d0;ad); 38
( ) = ( ) = {,,, } β ( ), < 1 ( ) + ( ) = ( ) + ( )
{ } ( ) = ( ) = {,,, } ( ) β ( ), < 1 ( ) + ( ) = ( ) + ( ) max, ( ) [ ( )] + ( ) [ ( )], [ ( )] [ ( )] = =, ( ) = ( ) = 0 ( ) = ( ) ( ) ( ) =, ( ), ( ) =, ( ), ( ). ln ( ) = ln ( ). + 1 ( ) = ( ) Ω[ (
LoadBalancingforMinimizingExecutionTimeofaTargetJobon anetworkofheterogeneousworkstations DepartmentofElectricalandComputerEngineering S.-Y.LeeandC.-H.Cho [email protected] Auburn,AL36849. AuburnUniversity
M.A. in Communication Studies
1 M.A. in Communication Studies Communication Studies Courses CM 7390 Seminar in Professional Development 3 CM 8300 Foundations in Communication Studies 3 CM 9300 Foundations in Communication Theory 3
Polynomial Factoring. Ramesh Hariharan
Polynomial Factoring Ramesh Hariharan The Problem Factoring Polynomials overs Integers Factorization is unique (why?) (x^2 + 5x +6) (x+2)(x+3) Time: Polynomial in degree A Related Problem Factoring Integers
recursively enumerable languages context-free languages regular languages
CPS 140 - Mathematical Foundations of CS Dr. S. Rodger Section: Recursively Enumerable Languages èhandoutè Deænition: A language L is recursively enumerable if there exists a TM M such that L=LèMè. if
OPTIMAL SELECTION BASED ON RELATIVE RANK* (the "Secretary Problem")
OPTIMAL SELECTION BASED ON RELATIVE RANK* (the "Secretary Problem") BY Y. S. CHOW, S. MORIGUTI, H. ROBBINS AND S. M. SAMUELS ABSTRACT n rankable persons appear sequentially in random order. At the ith
Solutions of Equations in One Variable. Fixed-Point Iteration II
Solutions of Equations in One Variable Fixed-Point Iteration II Numerical Analysis (9th Edition) R L Burden & J D Faires Beamer Presentation Slides prepared by John Carroll Dublin City University c 2011
CODES FOR PHARMACY ONLINE CLAIMS PROCESSING
S FOR PHARMACY ONLINE CLAIMS PROCESSING The following is a list of error and warning codes that may appear when processing claims on the online system. The error codes are bolded. CODE AA AB AI AR CB CD
An Incomplete C++ Primer. University of Wyoming MA 5310
An Incomplete C++ Primer University of Wyoming MA 5310 Professor Craig C. Douglas http://www.mgnet.org/~douglas/classes/na-sc/notes/c++primer.pdf C++ is a legacy programming language, as is other languages
On closed-form solutions of a resource allocation problem in parallel funding of R&D projects
Operations Research Letters 27 (2000) 229 234 www.elsevier.com/locate/dsw On closed-form solutions of a resource allocation problem in parallel funding of R&D proects Ulku Gurler, Mustafa. C. Pnar, Mohamed
CSE 135: Introduction to Theory of Computation Decidability and Recognizability
CSE 135: Introduction to Theory of Computation Decidability and Recognizability Sungjin Im University of California, Merced 04-28, 30-2014 High-Level Descriptions of Computation Instead of giving a Turing
DECLARATION OF PERFORMANCE NO. HU-DOP_TN-212-25_001
NO. HU-DOP_TN-212-25_001 Product type 212 (TN) 3,5x25 mm EN 14566:2008+A1:2009 NO. HU-DOP_TN-212-35_001 Product type 212 (TN) 3,5x35 mm EN 14566:2008+A1:2009 NO. HU-DOP_TN-212-45_001 Product type 212 (TN)
DECLARATION OF PERFORMANCE NO. HU-DOP_TD-25_001
NO. HU-DOP_TD-25_001 Product type TD 3,5x25 mm EN 14566:2008+A1:2009 NO. HU-DOP_TD-35_001 Product type TD 3,5x35 mm EN 14566:2008+A1:2009 NO. HU-DOP_TD-45_001 Product type TD 3,5x45 mm EN 14566:2008+A1:2009
Statistical Machine Translation: IBM Models 1 and 2
Statistical Machine Translation: IBM Models 1 and 2 Michael Collins 1 Introduction The next few lectures of the course will be focused on machine translation, and in particular on statistical machine translation
New Investigator Form
Research Grants Management System User Guide New Investigator Application Form Target Audience 2016 NHMRC Project Grants New Investigators and/or Cancer Australia Young Investigators 9 December 2015 1
Stock Exchange of Mauritius Ground Rules for the SEM-10
Stock Exchange of Mauritius Ground Rules for the SEM-10 The Stock Exchange of Mauritius is introducing a new index, the SEM-10 Index, comprising shares listed on the Official Market. Designed to meet international
On the Eigenvalues of Integral Operators
Çanaya Üniversitesi Fen-Edebiyat Faültesi, Journal of Arts and Sciences Say : 6 / Aral 006 On the Eigenvalues of Integral Operators Yüsel SOYKAN Abstract In this paper, we obtain asymptotic estimates of
PINPOINT: What and Where?
604.879.4280 [email protected] 2014 Mi ki- B.C. Ki Ly Ifm Smb 2009 BRITISH COLUMBIA EDITION Fb S BRITI H y 2010 PINPOINT:? Ii i I: Ti m v mmiz ci b fiv m ii c fm BCCA i A. A i f c, A i m f im icki, v b c i. ITION
I. GROUPS: BASIC DEFINITIONS AND EXAMPLES
I GROUPS: BASIC DEFINITIONS AND EXAMPLES Definition 1: An operation on a set G is a function : G G G Definition 2: A group is a set G which is equipped with an operation and a special element e G, called
Wes, Delaram, and Emily MA751. Exercise 4.5. 1 p(x; β) = [1 p(xi ; β)] = 1 p(x. y i [βx i ] log [1 + exp {βx i }].
Wes, Delaram, and Emily MA75 Exercise 4.5 Consider a two-class logistic regression problem with x R. Characterize the maximum-likelihood estimates of the slope and intercept parameter if the sample for
I I I I I I I I I I I I I I I I I I I
Abstract Decision Making with nterval nfluence Diagrams John S. Breese [email protected] Kenneth W. Fertig [email protected] Rockwell Science Center, Palo Alto Laboratory 444 High Street Palo Alto, CA 94301
BIG DATA PROBLEMS AND LARGE-SCALE OPTIMIZATION: A DISTRIBUTED ALGORITHM FOR MATRIX FACTORIZATION
BIG DATA PROBLEMS AND LARGE-SCALE OPTIMIZATION: A DISTRIBUTED ALGORITHM FOR MATRIX FACTORIZATION Ş. İlker Birbil Sabancı University Ali Taylan Cemgil 1, Hazal Koptagel 1, Figen Öztoprak 2, Umut Şimşekli
LIMITS AND CONTINUITY
LIMITS AND CONTINUITY 1 The concept of it Eample 11 Let f() = 2 4 Eamine the behavior of f() as approaches 2 2 Solution Let us compute some values of f() for close to 2, as in the tables below We see from
A Permutation Network
A Permutation Network ABRAHAM WAKSMAN Stanford Research nstitute, Menlo Park, California,~BS'rm~CT. n this paper the construction of a switching network capable of n!-permutation of its n is:put terminals
CLOUDS: A Decision Tree Classifier for Large Datasets
CLOUDS: A Decision Tree Classifier for Large Datasets Khaled Alsabti Department of EECS Syracuse University Sanjay Ranka Department of CISE University of Florida Vineet Singh Information Technology Lab
1.- L a m e j o r o p c ió n e s c l o na r e l d i s co ( s e e x p li c a r á d es p u é s ).
PROCEDIMIENTO DE RECUPERACION Y COPIAS DE SEGURIDAD DEL CORTAFUEGOS LINUX P ar a p od e r re c u p e ra r nu e s t r o c o rt a f u e go s an t e un d es a s t r e ( r ot u r a d e l di s c o o d e l a
Option Pricing. Chapter 12 - Local volatility models - Stefan Ankirchner. University of Bonn. last update: 13th January 2014
Option Pricing Chapter 12 - Local volatility models - Stefan Ankirchner University of Bonn last update: 13th January 2014 Stefan Ankirchner Option Pricing 1 Agenda The volatility surface Local volatility
Modern Physics Laboratory e/m with Teltron Deflection Tube
Modern Physics Laboratory e/m with Teltron Deflection Tube Josh Diamond & John Cummings Fall 2010 Abstract The deflection of an electron beam by electric and magnetic fields is observed, and the charge
Rain Sensor "AWS" TYPE CHART and INSTALLATION INSTRUCTION
Typ:.N. 9.N. 9 / /g / / / / / up P X 0/0/0, 9/9 0/9 y pug, 09/90/0 / / / / ( L/U 0 ) /g / / P X /09/0, fm P p, 09/9 0/00 / / / / ( D ) /p / /y P X 9/09/0, fm B / E p, p //0, 09/90/0 / / / / (C / B ) /p
Overview of Number Theory Basics. Divisibility
Overview of Number Theory Basics Murat Kantarcioglu Based on Prof. Ninghui Li s Slides Divisibility Definition Given integers a and b, b 0, b divides a (denoted b a) if integer c, s.t. a = cb. b is called
IRGP4068DPbF IRGP4068D-EPbF
INSULATED GATE BIPOLAR TRANSISTOR WITH ULTRA-LOW VF DIODE FOR INDUCTION HEATING AND SOFT SWITCHING APPLICATIONS Features Low V CE (ON) Trench IGBT Technology Low Switching Losses Maximum Junction temperature
Manual for SOA Exam MLC.
Chapter 5. Life annuities. Extract from: Arcones Manual for the SOA Exam MLC. Spring 2010 Edition. available at http://www.actexmadriver.com/ 1/114 Whole life annuity A whole life annuity is a series of
OHJ-2306 Introduction to Theoretical Computer Science, Fall 2012 8.11.2012
276 The P vs. NP problem is a major unsolved problem in computer science It is one of the seven Millennium Prize Problems selected by the Clay Mathematics Institute to carry a $ 1,000,000 prize for the
An Introduction to the RSA Encryption Method
April 17, 2012 Outline 1 History 2 3 4 5 History RSA stands for Rivest, Shamir, and Adelman, the last names of the designers It was first published in 1978 as one of the first public-key crytographic systems
Monolithic Amplifier PMA2-43LN+ Ultra Low Noise, High IP3. 50Ω 1.1 to 4.0 GHz. The Big Deal
Ultra Low Noise, High IP3 Monolithic Amplifier 50Ω 1.1 to 4.0 GHz The Big Deal Ultra low noise figure, 0.46 db High gain, high IP3 Small size, 2 x 2 x 1mm 2mm x 2mm Product Overview Mini-Circuits is an
MAT 1341: REVIEW II SANGHOON BAEK
MAT 1341: REVIEW II SANGHOON BAEK 1. Projections and Cross Product 1.1. Projections. Definition 1.1. Given a vector u, the rectangular (or perpendicular or orthogonal) components are two vectors u 1 and
Optimal order placement in a limit order book. Adrien de Larrard and Xin Guo. Laboratoire de Probabilités, Univ Paris VI & UC Berkeley
Optimal order placement in a limit order book Laboratoire de Probabilités, Univ Paris VI & UC Berkeley Outline 1 Background: Algorithm trading in different time scales 2 Some note on optimal execution
Chapter 7. Continuity
Chapter 7 Continuity There are many processes and eects that depends on certain set of variables in such a way that a small change in these variables acts as small change in the process. Changes of this
Social Media Mining. Data Mining Essentials
Introduction Data production rate has been increased dramatically (Big Data) and we are able store much more data than before E.g., purchase data, social media data, mobile phone data Businesses and customers
Introduction to Mobile Robotics Bayes Filter Particle Filter and Monte Carlo Localization
Introduction to Mobile Robotics Bayes Filter Particle Filter and Monte Carlo Localization Wolfram Burgard, Maren Bennewitz, Diego Tipaldi, Luciano Spinello 1 Motivation Recall: Discrete filter Discretize
ERDOS PROBLEMS ON IRREGULARITIES OF LINE SIZES AND POINT DEGREES A. GYARFAS*
BOLYAI SOCIETY MATHEMATICAL STUDIES, 11 Paul Erdos and his Mathematics. II, Budapest, 2002, pp. 367-373. ERDOS PROBLEMS ON IRREGULARITIES OF LINE SIZES AND POINT DEGREES A. GYARFAS* Problems and results
Statistical Machine Learning
Statistical Machine Learning UoC Stats 37700, Winter quarter Lecture 4: classical linear and quadratic discriminants. 1 / 25 Linear separation For two classes in R d : simple idea: separate the classes
Determination of the normalization level of database schemas through equivalence classes of attributes
Computer Science Journal of Moldova, vol.17, no.2(50), 2009 Determination of the normalization level of database schemas through equivalence classes of attributes Cotelea Vitalie Abstract In this paper,
The Goldberg Rao Algorithm for the Maximum Flow Problem
The Goldberg Rao Algorithm for the Maximum Flow Problem COS 528 class notes October 18, 2006 Scribe: Dávid Papp Main idea: use of the blocking flow paradigm to achieve essentially O(min{m 2/3, n 1/2 }
www.fmtalkinghouse.com
www.fmtalkinghouse.com 24x7 FM Announcement System (Talking Sign) 2009 FM TALKING HOUSE All rights reserved Talking Sign 24x7 FM Announcement System Description: The 24x7 FM Announcement System or Talking
Measuring evolution of systemic risk across insurance-reinsurance company networks. Abstract
Measuring evolution of systemic risk across insurance-reinsurance company networks Abstract Constructing a stochastic model to quantify counterparty risk in the insurance industry Create estimators based
4.6 Linear Programming duality
4.6 Linear Programming duality To any minimization (maximization) LP we can associate a closely related maximization (minimization) LP. Different spaces and objective functions but in general same optimal
Oscar E. Morel UtilX Corporation
Oscar E. Morel UtilX Corporation Time Domain Reflectometry (TDR) has been the preferred technique to assess: Cable length Splice number and spatial location, and Metallic neutral condition Tests for neutral
Eris Interest Rate Swap Futures: Flex Contract Specifications
Eris Interest Rate Swap Futures: Flex Contract Specifications Trading Hours Contract Structure Contract Size Trading Conventions Swap Futures Leg Conventions Effective Date Cash Flow Alignment Date ( CFAD
B I N G O B I N G O. Hf Cd Na Nb Lr. I Fl Fr Mo Si. Ho Bi Ce Eu Ac. Md Co P Pa Tc. Uut Rh K N. Sb At Md H. Bh Cm H Bi Es. Mo Uus Lu P F.
Hf Cd Na Nb Lr Ho Bi Ce u Ac I Fl Fr Mo i Md Co P Pa Tc Uut Rh K N Dy Cl N Am b At Md H Y Bh Cm H Bi s Mo Uus Lu P F Cu Ar Ag Mg K Thomas Jefferson National Accelerator Facility - Office of cience ducation
Some Essential Statistics The Lure of Statistics
Some Essential Statistics The Lure of Statistics Data Mining Techniques, by M.J.A. Berry and G.S Linoff, 2004 Statistics vs. Data Mining..lie, damn lie, and statistics mining data to support preconceived
GREATEST COMMON DIVISOR
DEFINITION: GREATEST COMMON DIVISOR The greatest common divisor (gcd) of a and b, denoted by (a, b), is the largest common divisor of integers a and b. THEOREM: If a and b are nonzero integers, then their
Large-Scale Data Sets Clustering Based on MapReduce and Hadoop
Journal of Computational Information Systems 7: 16 (2011) 5956-5963 Available at http://www.jofcis.com Large-Scale Data Sets Clustering Based on MapReduce and Hadoop Ping ZHOU, Jingsheng LEI, Wenjun YE
How To Solve A Sequential Mca Problem
Monte Carlo-based statistical methods (MASM11/FMS091) Jimmy Olsson Centre for Mathematical Sciences Lund University, Sweden Lecture 6 Sequential Monte Carlo methods II February 3, 2012 Changes in HA1 Problem
Trimming a Tree and the Two-Sided Skorohod Reflection
ALEA, Lat. Am. J. Probab. Math. Stat. 12 (2), 811 834 (2015) Trimming a Tree and the Two-Sided Skorohod Reflection Emmanuel Schertzer UPMC Univ. Paris 06, Laboratoire de Probabilités et Modèles Aléatoires,
On continued fractions of the square root of prime numbers
On continued fractions of the square root of prime numbers Alexandra Ioana Gliga March 17, 2006 Nota Bene: Conjecture 5.2 of the numerical results at the end of this paper was not correctly derived from
IFANCA HALAL PRODUCT CERTIFICATE
Certificate No.:PUR.6106.6107.150016.US Page 1 of 5 Manufactured at : Gradings: E, S, FL, C, MB, CL, TL, PL, DL,PF,PP,PLUS, LT, EDK, MR, FG Ionic Forms: Na, H, K, Ca 1. Purolite C100 2. Purolite C100X10
Christfried Webers. Canberra February June 2015
c Statistical Group and College of Engineering and Computer Science Canberra February June (Many figures from C. M. Bishop, "Pattern Recognition and ") 1of 829 c Part VIII Linear Classification 2 Logistic
Frsq: A Binary Image Coding Method
Frsq: A Binary Image Coding Method Peter L. Stanchev, William I. Grosky, John G. Geske Kettering University, Flint, MI 4854, {pstanche, jgeske}@kettering.edu University of Michigan-Dearborn, Dearborn,
9th Max-Planck Advanced Course on the Foundations of Computer Science (ADFOCS) Primal-Dual Algorithms for Online Optimization: Lecture 1
9th Max-Planck Advanced Course on the Foundations of Computer Science (ADFOCS) Primal-Dual Algorithms for Online Optimization: Lecture 1 Seffi Naor Computer Science Dept. Technion Haifa, Israel Introduction
FM 55-30 14 APRIL 2000 By Order of the Secretary of the Army: Official: ERIC K. SHINSEKI General, United States Army Chief of Staff Administrative Assistant to the Secretary of the Army 0005503 DISTRIBUTION:
Comm. Korean Math. Soc. 13 (1998), No. 4, pp. 913{931 INVARIANT CUBATURE FORMULAS OVER A UNIT CUBE Kyoung Joong Kim and Man Suk Song Abstract. Using invariant theory, new invariant cubature formulas over
Using the Normalized Image Log-Slope, part 3
T h e L i t h o g r a p h y E x p e r t (Summer 2001) Using the Normalized Image Log-Slope, part 3 Chris A. Mack, FINLE Technologies, A Division of KLA-Tencor, Austin, Texas As we saw in parts 1 and 2
Chapter 13: Basic ring theory
Chapter 3: Basic ring theory Matthew Macauley Department of Mathematical Sciences Clemson University http://www.math.clemson.edu/~macaule/ Math 42, Spring 24 M. Macauley (Clemson) Chapter 3: Basic ring
Today s Topics. Primes & Greatest Common Divisors
Today s Topics Primes & Greatest Common Divisors Prime representations Important theorems about primality Greatest Common Divisors Least Common Multiples Euclid s algorithm Once and for all, what are prime
Correlation. Alan T. Arnholt Department of Mathematical Sciences Appalachian State University [email protected]
Correlation Alan T. Arnholt Department of Mathematical Sciences Appalachian State University [email protected] Spring 2006 R Notes 1 Copyright c 2006 Alan T. Arnholt 2 Correlation Overview of Correlation
Applied Algorithm Design Lecture 5
Applied Algorithm Design Lecture 5 Pietro Michiardi Eurecom Pietro Michiardi (Eurecom) Applied Algorithm Design Lecture 5 1 / 86 Approximation Algorithms Pietro Michiardi (Eurecom) Applied Algorithm Design
Journal of Chemical and Pharmaceutical Research, 2014, 6(3):34-39. Research Article. Analysis of results of CET 4 & CET 6 Based on AHP
Available online www.jocpr.com Journal of Chemical and Pharmaceutical Research, 2014, 6(3):34-39 Research Article ISSN : 0975-7384 CODEN(USA) : JCPRC5 Analysis of results of CET 4 & CET 6 Based on AHP
TIRANTS ARTITEC INOX TARIF BRUT (HTVA)
ART.PH60.01 TIRANT INOX PH60 forme U / 19X125 mm + Fix.travers.décoratif 23,94 ART.PH60.02 TIRANT INOX PH60 forme U / 19X125 mm + VIS TF M8 15,96 ART.PH60.03 TIRANT INOX PH60 forme U / 19X125 mm + FIX.
Foundations of Machine Learning On-Line Learning. Mehryar Mohri Courant Institute and Google Research [email protected]
Foundations of Machine Learning On-Line Learning Mehryar Mohri Courant Institute and Google Research [email protected] Motivation PAC learning: distribution fixed over time (training and test). IID assumption.
CIPFA. Interactive Timetable. 2016 Live Online
CIPFA Interactive Timetable 2016 Live Online Version 5: Information last updated 09/08/16 Please note: Information and dates in this timetable are subject to change CIPFA Information 2016 The following
MA4001 Engineering Mathematics 1 Lecture 10 Limits and Continuity
MA4001 Engineering Mathematics 1 Lecture 10 Limits and Dr. Sarah Mitchell Autumn 2014 Infinite limits If f(x) grows arbitrarily large as x a we say that f(x) has an infinite limit. Example: f(x) = 1 x
Big Data trifft Industrie Im Internet der Bosch-Dinge und -Dienste
Big Data trifft Industrie Im Internet der Bosch-Dinge und -Dienste Dr. Lothar Baum Februar 2015 1 Bosch Business Areas Automotive Technology Industrial Technology Energy and Building Technology Consumer
HCC4541B HCF4541B PROGRAMMABLE TIMER
HCC4541B HCF4541B PROGRAMMABLE TIMER 16 STAGE BINARI COUNTER LOW SYMMETRICAL OUTPUT RESISTANCE, TYPICALLY 100 OHM AT DD = 15 OSCILLATOR FREQUENCY RANGE : DC TO 100kHz AUTO OR MASTER RESET DISABLES OSCIL-
Breaking The Code. Ryan Lowe. Ryan Lowe is currently a Ball State senior with a double major in Computer Science and Mathematics and
Breaking The Code Ryan Lowe Ryan Lowe is currently a Ball State senior with a double major in Computer Science and Mathematics and a minor in Applied Physics. As a sophomore, he took an independent study
Representation of functions as power series
Representation of functions as power series Dr. Philippe B. Laval Kennesaw State University November 9, 008 Abstract This document is a summary of the theory and techniques used to represent functions
Amply Fws Modules (Aflk n, Sonlu Zay f Eklenmifl Modüller)
Çankaya Üniversitesi Fen-Edebiyat Fakültesi, Journal of Arts and Sciences Say : 4 / Aral k 2005 Amply Fws Modules (Aflk n, Sonlu Zay f Eklenmifl Modüller) Gökhan B LHA * Abstract In this work amply finitely
General Specifications
General Specifications ProSafe-RS Outline of I/O Modules [Release 3] GENERAL This GS describes the specifications of I/O modules to be mounted on the safety control unit or the safety node unit of the
Artificial Neural Networks and Support Vector Machines. CS 486/686: Introduction to Artificial Intelligence
Artificial Neural Networks and Support Vector Machines CS 486/686: Introduction to Artificial Intelligence 1 Outline What is a Neural Network? - Perceptron learners - Multi-layer networks What is a Support
MapReduce and Distributed Data Analysis. Sergei Vassilvitskii Google Research
MapReduce and Distributed Data Analysis Google Research 1 Dealing With Massive Data 2 2 Dealing With Massive Data Polynomial Memory Sublinear RAM Sketches External Memory Property Testing 3 3 Dealing With
Monte Carlo-based statistical methods (MASM11/FMS091)
Monte Carlo-based statistical methods (MASM11/FMS091) Magnus Wiktorsson Centre for Mathematical Sciences Lund University, Sweden Lecture 6 Sequential Monte Carlo methods II February 7, 2014 M. Wiktorsson
Universal hashing. In other words, the probability of a collision for two different keys x and y given a hash function randomly chosen from H is 1/m.
Universal hashing No matter how we choose our hash function, it is always possible to devise a set of keys that will hash to the same slot, making the hash scheme perform poorly. To circumvent this, we
PROMPT ENGINEERING & TRADING SERVICES CO. W.L.L P.O. BOX 24067 BARWA VILLAGE, WAKRA, BUILDING NO. 6 SHOP NO. 19 AND 20 DOHA QATAR
This is to signify that PROMPT ENGINEERING & TRADING SERVICES CO. W.L.L P.O. BOX 24067 BARWA VILLAGE, WAKRA, BUILDING NO. 6 SHOP NO. 19 AND 20 DOHA QATAR Calibration Laboratory CL-165 has met the requirements
Quick Disconnect. Wiring diagrams show quick. Male Receptacle End View Blue. disconnect pin numbers ( ) 1 Brown. Load. 4 1 Black ( ) 3 Blue ( ) Brown
Pile Driver Sensor One-piece stainless steel cylindrical housing 2-wire DC 3-wire DC 2-wire AC/DC 1. to sensing range dia. model: The Die Guy TM Rugged Solid stainless steel housing withstands abusive
Real-time Targeted Influence Maximization for Online Advertisements
Real-time Targeted Influence Maximization for Online Advertisements Yuchen Li Dongxiang Zhang ian-lee Tan Department of Computer Science School of Computing, National University of Singapore {liyuchen,zhangdo,tankl}@comp.nus.edu.sg
INSTALLATION OPERATING & MAINTENANCE MANUAL
INSTALLATION OPERATING & MAINTENANCE MANUAL Basic Climatic TM Controller Ecologic English November 2002 CONTENTS PAGE GENERAL DESCRIPTION 3 USER INTERFACE 4 The keypad incorporated on the unit 4 The keypad
