資源描述：
《continuous distributionsand the poisson process》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在工程資料-天天文庫(kù)。
1、CHAPTEREIGHTContinuousDistributionsandthePoissonProcessThischapterintroducesthegeneralconceptofcontinuousrandomvariables,focusingontwoexamplesofcontinuousdistributions:theuniformdistributionandtheexponentialdistribution.WethenproceedtostudythePoissonprocess,acontinuoustimecountingprocessthatisrel
2、atedtoboththeuniformandexponentialdistributions.WeconcludethischapterwithbasicapplicationsofthePoissonprocessinqueueingtheory.8.1.ContinuousRandomVariables8.1.1.ProbabilityDistributionsinRThecontinuousroulettewheelinFigure8.IhascircumferenceI.Wespinthewheel,andwhenitstops.theoutcomeistheclockwise
3、distanceX(computedwithinfiniteprecision)fromthe··o··marktothearrow.ThesamplespaceQofthisexperimentconsistsofallrealnumbersintherange[0,I).Assumethatanypointonthecircumferenceofthediskisequallylikelytofacethearrowwhenthediskstops.Whatistheprobabilitypofagivenoutcomex'?Toanswerthisquestion,werecall
4、thatinChapterIwedefinedaprobabilityfunctiontobeanyfunctionthatsatisfiesthefollowingthreerequirements:1.Pr(Q)=I;2.foranyeventE,0.SPr(E).SI;3.forany(finiteorenumerable)collectionBofdisjointevents,Pr(ur;)=LPr(f:).EEBEEBLetS(k)beasetofkdistinctpointsintherange[0,1),andletpbetheprobabilitythatanygiven
5、pointin[0,I)istheoutcomeoftherouletteexperiment.SincetheprobabilityofanyeventisboundedbyI,1888.1CONTINUOUSRANDOMVARIABLESFigure8.1:Acontinuousroulettewheel.Pr(xES(k))=kpS1.\·ecanchooseanynumberkofdistinctpointsintherange[0,I).sowemusthave?/)sIforanyintegerk,whichimpliesthatp=0.Thus,weobservethat
6、inaninfinite'amplespacetheremaybepossibleeventsthathaveprobability0.Takingthecomplementofsuchanevent,weobservethatinaninfinitesamplespacetherecanbeevents\ithprobabilityIthatdonotcorrespondtoallpossibleexperimentaloutcomes.andthustherecanbeeventswithprobabilityIthatare,insomesense.notcertain?lfth
7、eprobabilityofeachpossibleoutcomeofourexperimentis0.howdowedefinetheprobabilityoflargereventswithnonzeroprobability'?Forprobabilitydistributions,,erR,probabilitiesareassignedtointen·aisratherthantoindiidual?tlue:-;.
