I DiscreteTime part I continuous Time Markov chain An example of discrete time Markov chain Definition Branching Process Branching Process discretetime Markov chain are i.i.cl random variables withpmf SM 9 is and 加⼆ a niiiiiiiii nslo.tn 知⼼ 9i i nth generation in mi Xuemeans thenumbers of individuals number of children of ith individual Fort n th generation 9_9means the theextinctionprobability Main question Howto compute let 419 Ě Pk Ok o egg Theorem extinction P 如0 forsome n Then 0 let up O 0 Egg u is thesmallest rut of 419 extinctbythe nth generation P Mio lintmybytthgy Let an P Un 4 Uni To and u compute u or un find Xu 0 you nz can and 4191 find thesmallest not 419 find Pk Ko 1,2 fu Assignment solve un 4 question l 2 undum iteratively withinitialcondition u Exercise 1 Find u and Us Pia Rita let pmffft 䧇V 1 a 90 o 94a. i Ok ĒR 419 a ag ou solvedD 9 l ag2 a 7ag2 O l1 a Aia 灬 a win 兴 I U 1 If a u pl U u 1 2 1 兴 1 92 0 1 哥 extinct by the third generation 0 a Un a f 4M 1 410 411 a Uz flu 10U2 f Um 以 u 以 x 1 呰 ocaci.IE If U3 吢 81 0 a a a 04 t a a 1 1 a 4 ta a12 2 1 11 a 9 ta 1 a 4zacl.ci tll a d11 a3ll al4tza4taP4a taR ctaj q a 4 1 Time Markov Chain and Poisson Process continuous Time MarkovChain Definition Markov chain is a stochastic process A continuous time 1,2 index set To too and theState space Sc go Dare I continuous SXHYuit .tk that satisfies the Markov property t.tn 1 U of t.octic.tc.nc.tn in l XHnkin X Gm Kim Plan 㗊 vi i Main Main X 01 i P 蕊 燕 苎 蕊 in ES i An importantant Before introduce class of CTMC Poisson Process we need to know two types of distribution Poisson Process Poisson distribution variable A discrete random Definition X is said to follow 1,2 Kk etf KEfo with parameter µso if P a If XvPoisson W then Et Var X Duperty If Xv Poissonlui Y v PoissonWH Xi Y then Xt Yu Poisson Mtn a Poissondistr.hn 了 lb Exponential distribution continuous random variable Definition A with rate ⼊ if PT f Var ㄒ 六 ffkofmnremy.JP the distribution of does we depend on how ii has exponential distribution eit.tt t Property O ET T Pts IT watt times men b PT t to an has passed d l s event dusider a post office that Exercise 2 MA run by two discovers that enters theoffice and or Mr.B is beingservedby Mr.A willbegin his service as soon as If theamount of time that a clerkspends clerkland Mv.B.by clerk2 Mr.B clerks Mr.Bz leaves distributed with rate ⼊ with a customer is exponentially that ofthethreecustomers Mr.A is the last probability the Find leave thepose office 想 辔 n_n 1 Let let let 2 if Bzleave first.tn 䃕 䎕 花 begin his service A that t be the time exponan.dk T service time the A's T.be cho are of customer th time T be the remaining still being served at the 27 ⼀⼀ iiipi.tt Luhy to the service tire let TB 有 be If 131 leave first leavefirst P 估 S 1131 If 137 leave first A Bi 132 P 在 234t ⽯ t P 下32 75 P 历 175 P Tzss1132 leave 却傆 ÍP 佦 Lau of total prob i 所以 P 有275 左 有2 左 expo ⼋ ˋ Poisson Process Poisson Process Definition A homogeneous X Poisson process X of rate ⼊ is ⼆0 f any time points to ⼼ ctu a GMC thatsatisfies theprocess increments Xltn Xltm are independent XG.tl lto Xltz X Poisson at X Stt X s so and t so for any time events that occur by number of Xt means the total ti Theimǜǚìtsfi then To下 ⽯ are then and N ii.cl and Tine exponential its t event Exercise3 ⼋ 2 13 n Xlt is a Poisson process of rate ⼊ Pr 如 61 姙 8 E It T T ⼗万 Mi 2 If E X14⼈ 如 K find 2 3 8 B1 加 1 6M41 ⼆ X141 81 Pr 加1 6 Tg iii 不如 ⼼ 1 器 如 只炎 华 加 1 X141 P 4 ㄨ4 学 灐管 127 P 1 E To ⼗下 坧⼗万 砹繭 Th 1X ˋ 㖄 ⼊ 年 for2 HEKine more events 2 1 ⼤ 2 1 Ì 2 2 Elytra1 El X14__ 加 7 凹 巧 Poisson 2 ⼊ 2 ⼋ ⼆ 4 Exercise 4 A bus will arrive at a train The arrivaltime follows uniform1011 according to Passengers arrive Let X a station hours Tisson process with rat 7perhour thenumber ofpeople who geton the train 品品 台 Find ylbusarnle n la Y be the time that the train Then You U o 1 Xiu Poisson Tg Given Ky EM 的 7yvarlxlkgj arrive 7 EKEEMYKEMYK7 EE I Var K EN ar MY EMYD Var17Y t Var E17Y t 7 EY 1 49 Vary 49 ⾆ ⼆下 之 台