4.1 BINOMIAL DISTRIBUTION This is a discrete distribution. Suppose an experiment is conducted under essentially identical condi- Hons n times where these are independent trials. Let ‘p’ be the probability of occurrence of an even galled the probability of success and ‘q" its complementary event is called the probability of failure Then the probability of getting exactly r successes in n independent trials is nC,p’q""”. Definition A random variable has a Binomial distribution if X assumes only non-negative values and its prob- ability distribution is given by OL. , otherwise nC,p'q’,r PX --{ a) Caution. * nand p are called the parameters and hence we use the notation X ~ B(n, p) to say that X follows Binomial distribution with n and p as parameters. Scanned with CamSczStandard Discrete Distributions The assumptions followed by a binomial variate are as follows: There are only two possible outcomes for each trial called success and failure. The probability of success in each trial is constant. + The number of trials are independent. The probability distribution given by equation (4.1) is a probability mass function. IP(X =x Enc,p'a™ =nC,p q” +nC,p'q"" +..nC,p"q"” =(p+q)" =1 LPUX =x)=1 oe _ sD an-2(e-2) =] © EnC,pig’* = La-IC, pig = En 2C, Pa Scanned with CamSczMoments of Binomial Distribution Mean of Binomial Distribution E(X)= Ex P(X =x) =Lxnc pig ia n! _ xa—ay? progr) = tn1-@ oh =MEqic prigri-d E(X)=np an E(X*) = Ex°nC,p*q"* = L[x(x - 1) +x)nC,p*q"™ = Dx(x—nC,p'q"™ + LxnC, pq" noe xian?" == 2) eaten, 2 Sree (n-2)! -2 n-2-(x-2) @Din-2-~-ay 7? =Lx(x-1), =n(n—Dp? = E(X*)=n(n—N)p? +p Scanned with CamScz8 Probability, Statistics and Random Processes Variance of Binomial Distribution The variance of a distribution is given by V(X) = E(X?)-[ EP = n(n-l)p?+ np-n'p? = np? np? +np—n'p? =np(l-p) =npq Caution. ‘The mean of Binomial distribution is np and variance of Binomial distribution is npg. + The number of trials ‘nis finite. The problems relating to tossing of a coin or throwing a die or drawing cards from a packo! cards with replacement lead to Binomial dis istribution. Scanned with CamSczWorked Out Examples EXAMPLE 4.1 In a certain district, the need for money to buy drugs is stated as the reason for 75% of all thefts. Find the probability that among the next 5 theft cases reported in this district, (i) Exactly 2 resulted from the need for money to buy drugs. (ii) At most 3 resulted from the need for money to buy drugs Solution: Let denote the need for money to buy drugs in a certain district. Let p denote the percent- age of thefts in the city. p=0.75, -P —0.75 0.25 This represents a Binomial distribution. P(X =x) = Probability that there are exactly x successes =nC,p'q (i) P(X = 2) = Probability that exactly 2 resulted from the need for money to buy drugs 5C,p'q?* (0.75)? (0.25)? = 0.0878 (ii) P(X $3) = Probability that at most 3 resulted from the need for money to buy drugs Scanned with CamSczStandard Discrete Distributions © 119 PU S3)= PUY =0)4 PX =) 4 LX 2 2)4 PIX =3) SSCP + SCp'g 45C,pig! = (0.25)° +5 (0.75) ( = 0.0009765 +0.014648 = 0.367186 *45C,p'g™ Y* + 10(0.75)°(0.25)' + 10(0.75)3(0.25)? +0.08789 + 0.263671 EXAMPLE 4.2 Six cards are drawn from a pack of $2 cards. Find the probability that there are (i) At least three diamonds (i) None is a diamond (INTU Aug./Sep. 2008 set-1) Solution: Let V’denote a diamond card from the 52 pack of cards. PUX=x) Probability that there are exactly x successes =nC pq" (@)_p=Probability of drawing a diamond = 1. 13 q=l-p P(X 23) = Probability that there are at least three diamonds P(X 23)=1-P(X <3) (X =0)+ P(X =1) + P(X =2) | = 6C,p"g"* + 6Cp'g® +6C,p'g** 12) 1)(12)° (3)(2) ={= —|{=] +15] —] [= (2) +6(5\(8) 13) lia | = 0. 992377 Gi) Probability that none is a diamond = P(X = 0) = 0.618625 EXAMPLE 43 ; ; Two dice are thrown five times. If getting a double is a success, find the probability that getting the success (i) Ath ) Atleast once (INTU Aug//Sep. 2008, set-3) (i) Twice OCANNEU WIL UCallloCe120 Probability, Statistics and Random Processes Solution: Let Y denote the numbers on the two dice. Let p = getting a double when two dice are thr for getting a double = {(1, 1), (2,2), (3, 3), 4,4), (5, 5)s (6, 6)} 36 No. of favourable cas¢ Total number of cas oS ines Probability that there are exactly x successes =nC,p"q (@_PCX> 1) = Probability of getting success at least once P(X 21) =1-P(X <1) = 0.598122 (ii) P(X=2) = Probability of getting success exactly two times PU =2)=5C,p*q?* (J () = 0.160751 EXAMPLE 4.4 ut of 800 families with 5 children each, how many would you expect to have? (Three boys (ii) At least one boy (INTU Nov. 2006, se?) Solution: Let X denote the number of boys in a family. Probability of a boy = Probability of a girl 1 paa=5 P(X=x) = Probability that there are exactly x successes =nC,p*q"* (i) Probability of exactly 3 boys = P(X = 3) PX = s3 OCAMNEU WIL VallloCeStandard Discrete Distributions = 12] “(3 G) 272 = 10(0.03125) 3125 With 5 children each, the number of families with 3 boys each = Probability of exactly 3 boys x 800 = 800 (0.3125) = 250 Hence, out of 800 famili (ii) Probability of at least one boy = P(X > 1) P(X 21) P(X 21)=1-SC,p"q*” ““() 2 =1-0.03125 = 0.96875 EXAMPLE 4.5 ; a soning The mean of Binomial distribution is 3 and the variance is z Find the following: (i) Value of n (ii) P(X 27) ) PUSX<6) (INTU Aug. /Sep. 2008 set-4) (ii Solution: Given that the mean of Binomial distribution = np .2) np=3 42) In addition, given that the variance of Binomial distribution = "pq 1 (i) From equation (42), (3) na3x4e 12 OCdNNEU WILT Cdl NoCe122. © Probability, Statistics and Random Processes 10) + P(X =11)+ PX = 12) vy a Rye PX = 94 PX vy oT] PU = 14 PX = 8)+ PX i d Ge Dap roe iran pigs +12C,p°4" © 412Cgphg +12C pa" =12C.pq ACP . 412¢,p"q" ee tay 1Y (3) 1y°(3) ae 4»s(4} (3) +201) (3) +66{ +) 4 -m(1) (3) + Sala a)\4 a) \4 n'a) (1 (3) (3) () = 0.014253 P< X<6)= PIX = 1) +P = 2)4 PE =3)+ PU = 4)+ PX = 5) 2 12c pig?" +12C,p'g"? +12C,p'a" #12C,p'q?" +12C,p'q” : tray cay LV (3 ef 1¥(3Y" 4990/2 (3) (4) (3) en(t\(2 1) (3) 4200{ +] {=} +495 -n(4)(3) +66(1) (3) +220(2) (3 alla 1V(3) +72(2) (3) = 0.888111 EXAMPLE 4.6 1 The probability of a man hitting a target is = (i) Ifhe fires five times, what is the probability of his hitting the target at least twice? ii) How many times must he fire so that the probability of his hitting the target at least once is more than 90%? (INTU Aug/Sep. 2006 se!) Solution: Let n = number of trials = 5 p= probability of hitting a target = ; = probability of not hitting the target = (i) Probability of the man hitting the target at least twice = P(X > 2) P(X 22)=1-P(X <2) G16) -sE)G)] (3-3) OCaNNeEU WILT cai dbeStandard Discrete Distributions. = 123 ie “GI6) = 0.539095 (i) Given that probability of hitting the target at least once > 90% P (at least once) > 90% P(X 21) >90% 1-(&<1)>09 1-P(X=0)>0.9 Aan 1-ne,(3) (3) >09 (5) >09 The above equation is satisfied for n= 6 2¥ since (3) 0.9122 >0.9 EXAMPLE 4.7 In testing a certain kind of truck tyre over a rugged terrain, itis found that 25% of the trucks fail to complete the test run without a blowout and of the next 15 trucks tested, find the probability that: (From 3 to 6 have blowouts (ii) Fewer than 4 have blowouts (ii) More than 5 have blowouts Solution: Let n = Number of trucks tested = 15 robability of trucks failing to complete the test run g=1-p=0.75 This is a Binomial distribution. The probability distribution is given by. P(X =x)=nC,P"d P(X =x) = Probability that there are exactly x blowouts (Probability that there are between 3 and 6 trucks with Blowouls = PG 5) Pix<5)=1-P(X $5) — (P(X = 0) + POX = 1) + POX = 2) + PO = 3) + PK = 4) + P= 5} P(x >5)=1-{15C,(0.25)°(0.75)* +15C; (0.25)'(0.75)"* +1 5C,(0.25)°(0.75)" 415C,(0.25)°(0.75)" +15C, (0.25)*(0.75)" +15C, (0.25)*(0.75)" = 1 {0.013363+ 0.066817 + 0.155907 + 0.225199 + 0.168899 + 0.165146} =1-0.795331 = 0.204669 EXAMPLE 4.8 The mean and variance of a Binomial distribution are 4 and “respectively. Find P(X 21). 3 (INTU April/May. 2004, Solution: Mean of Binomial distribution = np Given that np =4 Variance of Binomial distribution = npq Cs) Given that npg = < a On dividing (4.4) by (4.3) we get, MPq np OCaNNEU WIL Lal oCeStandard Discrete Distributions 125 PIN 2 N= 1 PW et) =1- PX =0) PX =n=t-oo,(2) (If () = 0.998628 EXAMPLE 4.9 Ten coins are thrown simultaneously. Find the probability of getting at least seven heads. (INTU April/May. 2004 set -2) Solution: Let p = Probability of getting a head = [ele 4 = Probability of getting ahead = > The probability of getting x heads in a throw of 10 coins = P(X=.) 5) + ¥=0,1,2,...10 2 Probability of getting at least 7 heads when 10 coins are thrown = P(X 2 7) PUL 27) = P(X =7)+ PIX =8)+ P(X = 9)+ P(X =10) =19(3} (9) sme) @) me) eo GG) nic (3) se (ve) 6) -(y 10C, +10C, +10C, +10C.) (2) aaneasei0en 176 x 21 =2% = 0.1719 POE = F094 OCAMNEU WIL UalloceStatistics and Random Processes 126 Probability EXAMPLE 4.10 . | The probability that a man aged 75 years will die before his next birthday is 0.031. What isthe, at least 13 will reach their 76” birthday? (December'0s, tof 1S such men, ability Solution: Let p be the probability that a man aged 75 p=0031,¢=1-p=1—0.031 = 0.969 years will die before his next birthday. Number of men, n= 15 = Probability that out of 15 such men, at least 13 will reach their 76" birthday = P(X =13)+ P(X =14)+ P(X =15) = 15C,,(0.031)" (0.969)? +15C,, (0.031)"*(0.969)' +15C,,(0.031)'*(0.969)° = 2.40735E -18 +1.10022E — 20 +2.34653E — 23 PUX 21 = 2.41838E -18 = 2.418 x10" Scanned with CamSczFitting of Binomial Distribution A probability distribution is to be fit for a given set of data points. Here fitting of ‘Binomial distribute: is explained in detail. The main aim of fitting of a distribution is to find expected frequencies for a set data points (x, ,) which contain x, and corresponding observed frequencies f,, In order to find these ¥ find the probabilities of the Binomial distribution using the recurrence relation given in the next section Using these probabilities we find the expected frequencies, fox) = NP(x), where P(x) denotes the probabilities for x = 0,1,2,...n. N=S.f, where f, denote the observed frequencies of given data. Scanned with CamSczEXAMPLE 4.11 A set of 6 similar coins are tossed 640 times with the following results: a ads 0 1 Frequency 7 64 140 210 132 Be 8 Fit a Binomial distribution to the above data assuming that the coins are unbiased. Solution: The coin is unbiased: Probability of a head, p = | 2 Probability of a tail, g= | 2 Number of coins, n = 6 The recurrence relation for probabilities of Binomial distribution is D( = PC Ptl) (3) (x) The expected frequencies are calculated using f(x) = N- P(x) where N = 640 ro-s(3)(3) 8 (0) =N- POO) 1 me ray (S 5-2) pe) 6 P(I)=6 PO= Fy =10 s(y=N- PC) = 640 pa=(SS St)p oy Scanned with CamScz130 Probability. Statistics and Random Processes 56 15 »-> Pa Pay 2 a) 264 04 1s 2 N-PO 640) 150 ror mee nay 4 PQ) 3 £Q3)=N: PQ)= ovo =) = 200 P(a)= (S =). Pa) =37)= (2)-8 16) 64 15 (4)=N- P(4) = |= S(4) (4) ovo( ) 150 PS) = (<= “aa =2may=(2\(8) (23 nor-5reo=(3)(6)-(3) S()=N- P65) =640( }. 60 = 32 P(6)= (S a) 1 1\(3 1 P(6)=—P(s)=|—=|/ 2 )=(4 6") ( P(4) = 128(0.263393) = 33.7143 ~ 34 1-4 ’O=l a P Pa q =3P -(3 P(S)= 3 Fi P(4)= 3 (0.935205)(0.263393) = 0.147796 S(5) = N - P(5) =128(0.147796) = 18.91789 ~19 71-5 7@)=( 5) q 2 PO) == : P(S) a6 ) (0.935205)(0.147796) = 0.046073 S(6) = N - P(6) = 128(0.046073) = 5.897344 ~ 6 =(7-6)2 PC)= Teme) Ip 1 75PO= (3) (0.935205)(0.046073) = 0,006155 7 . OCANMNEU WIL UValloce P(?)=Standard Discrete Distributions = 135 f(T) = N+ P(7) = 128(0.006155) 0.78784 ~ 1 The expected frequencies of the data are as follows: ‘Number of heads 0 1 2 3 4 5 6 7 Total Frequency 7 6 19 35 (303 7 1 128 Expected Frequency, fix) 1 8 23 36 34 19 6 1 128 Scanned with CamScz4.2. POISSON DISTRIBUTION Definition Arandom variable X’is said to follow a Poisson distribution if it assumes only non-negative values and its probability distribution is given by, eye PW exnyelg TE023.- 0, otherwise Caution + Here, Zis called the parameter of the distribution. The above distribution is a probability mass function since eta x! Pug P(X =x) = Tio RoR eee Test gted A =1 Be a ee, Leo Gh = Eien Gap Gaal The following assumptions are made for Poisson distribution +n, the number of trials is very large, that is, n= +p, the constant probability of success for each trial is very small, that is, p> + np=A., is finite _ a following are some instances, where Poisson distribution may be successfully employed: + Number of printing mistakes at each page of a good book + Number of air accidents in some unit of time + Number of suicides reported in a particular city . + Number of defective material in a packing of a good conce Scanned with CamScz136 = Probability, Statistics and Random Processes Moments of Poisson Distribution Mean of Poisson Distribution If X is a poisson variate, mean of Vis given by, E(X) = Ti. X PX =*) . eta =LloXy ote @-D! “aye at = Bao yl eae BU)= AE Epi =e*e* E(X)=4 Variance of Poisson Distribution If-X is a poisson variate, variance of X is given by, W(X) = EX?) — {ECOY E(X*)=D7_)x°P(X =x) 2 =Ziot x! = Dio @-)+x] ety x! a =Er are) x! eta we + Elo = Xie xe 1) “+A, {since E(X) =A} V(X) = E(X*)- {EXP =R+A-2 OCANNEU WIL UCalloceStandard Discrete Distributions =" 137 Caution: Mean of poisson distribution = variance of poisson distribution = 2 Worked Out Examples EXAMPLE 4.13 Ifa poisson variate X’is such that P(x = 1) = 2 P(x = 2), find the following: @ P=) (ii) Mean (iii) Variance of X Solution: The probability distribution of Poisson distribution is given by, ngs | P(X =x)=© 4 | x! Given that P(x = 1) = 2 P(@e=2) etd eta? 1! 2! | 44 a) = 0.367879 (ii) Mean of Poisson distribution = (iii) Variance of Poisson distribution = A=1 EXAMPLE 4.14 1 “, A IFXis a poisson variate such that 3P(x = 4) => POr=2)+ P= 0), find the following: (i) Mean of X (ii) P(e <2) Solution: The probability distribution of Poisson dis P(X = xj= stribution is given by, ota x! 1 py=2)+ PO=0), Given that 3P(x = 4) 71 “Ags eth ete a o ALK 4 =—+1 ores] | OCANNEYU WIL UCalloce138 = Probability, Statistics and Random Processes A= +2, since A>0 (i) Mean of Poisson distribution = 4 = +2,-2 (ii) When 4=+2, P(x $2)=P(X = O)+ PX = 1) + P(X = 2) 4A + A+ PID 7 14244/2} set = 0.676676 Fitting of Poisson Distribution A probability distribution is to be fit for a given set of data points. Here, fitting of Poisson distribution is explained in detail. The main aim of fitting of a distribution is to find expected frequencies for a set of data points (x, f) which contain x, and corresponding observed frequencies f.. In order to find these, we find the probabilities of the Poisson distribution using the recurrence relation which is given in the next section. Using these probabilities, we find the expected frequencies /(x) = P(x), where P(x) denotes the probabilities forx=0, 1,2...n. N =f, where f denotes the observed frequencies of given data. Recurrence Relation of Probabilities of Poisson Distribution The probability distribution of Poisson distribution is PUX =x) = Plx)= (47) x! agen P(r +1)=24— 4.8) &+D= Cap . Dividing equations (4.2) by (4.1) we get, eooyel PX +x) _ (+d! Pe) ea x! Ax! a+)! an! ny +1) P(r+1)= (+= ED Px) » OCdi Neg WIL! CallloceEXAMPLE 4.15 Fit a ie distribution to the following data with respect to the number of red blood corpuscles (x) per cell: x 0 1 2 3 4 5 ‘Number of cells, f° 142 156 69 27 5 1 Solution: Mean of the Poisson distribution = Mean of the above distribution aot Zi 0142-41156 +2%69+3x2744x5+5x1 142+156+69+27+5+1 A= =1 Number of red corpuscles, n = 5 The recurrence relation for probabilities of Poisson distribution is, P(xtl)= (4 a ;) Pe The expected frequencies are calculated using, fix) =N P(x) where N= 400 sta on x! 0! = 0.367879 f()=N-P(O) = 400 (0.367879) = 147.1516 ~147 P= (a aa)Po P(i) = 0.367879 f(t) =N PC) = 400 (0.367879) =147.1516 ~147 P= (Gi a)ro P(0)= PQ) L 2 Scanned with CamScz140 = Probability, Statistics and Random Processes f(Q=N-P(2) = 400(0.18394) = 73.576 ~ 74 PQ)= “(a sy) PQ)= 5 PQ) 1 = (JJoass9 = 0.061313 fQ)=N-PQ) = 400(0.61313) = 24,5252 ~ 25 Po-(55 ;)Po) Peay=4 70) 1 = (0.061313) = 0.015328 1(4) =N- P(A) = 400(0.015328) =6.1312~6 PS)= (4 ma} al payi(t PG) = 3? =(2}o01s329) = 0.003066 4) =N -P(5) = 400(0.003066) = 1.2264 ~1 OCalineu witil CalnsceStandard Discrete Distributions © 14] ‘The expected frequencies of the given data are obtained as follows: x ai ; = = __?2 3 4 5 Total Number of cells, f 215669 ”SCUCOO Expected frequency. f(x) Wy we 6 1 400 EXAMPLE 4.16 Fit a Poisson distribution to the following data by calculating the expected frequencies: a 0 1 2 3 4 5 6 7 8 f fh ey oe ee fe Solution: Mean of the Poisson distribution = Mean of the above distribution xx,fi xf OTL 1124211743487 427 4 SLL 6S THLE Fa 1124117 +574 27411434141 _73 ~ 400 = 1.7825 n=8 ‘The recurrence relation for probabilities of Poisson distribution is PQ tl)= (4 Ayre “The expected frequencies are calculated using fla) = N- P(x), where N= 400 Aa = 400 (0.168217) = 67.2868 ~ 67 Pu= ee 12825) po Pad. .7825)(0.168217) 299847 Scanned with CamScz42 = Probability, Statistics and Random Processes sl) =N- PCI) = 400 (0.299847) = 119.9388 ~ 120 7825 PQ) -( Pay lel (2 782 °)ruy -( ; Le )oo. 299847) = 0.267239 f(2)=N-P(2) = 400(0.267239) = 106.8956 ~ 107 PQ) = (es Sp pay=! 1825 py (S aka Jo: 267239) = 0.158785 £@)=N +P) = 400(0.158785) = 63.514 ~ 64 P(4) = (ae i \r0) pay=l: 1125 1,78 = 15 (0 158785) = 0.070759 S(4)=N P(A) = 400(0.070759) = 28,3036 ~ 28 PS) = (2) Pts) = ERPS pay = (23 Joororsy = 0.025226 . canned wit CarnSceStandard Discrete Distributions © 143 f(5) = N -P(5) - 400(0.025226) = 10.0904 = 1 , 1.7825 1.7825 Ae 25 6 (1.7825 16) ES aes) ( ‘ Jroorsar6y = 0.007494 1 (6) = N- P(6) = 400(0.007494) = 2.9976 ~3 5) _ ( 1.7825 Pw = (es P6 ea } (6) 4) _ 1.782 p= : ¢6)=( 7825 oor = 0.001908 ra P(7) = 400(0.001908) = 0.7632 ~1 1.7825 = (17825 pq P) ( 741 ) ” 1,7825 7825 P(8)= P(7) =| ——— |(0.001908) 8 8 = 0.000425 (8) = N -P(8) = 400(0.000425) =0.17~0 i tained as follows: es of the given data are ol The expected frequencies of the 8 _ a ; 2 ; : 1 400 i f 7 12 nN7 37 : “ 67 120 107 64 28 10 fx) 0 A fi ere are on the average 2 F » proofreader finds that there are on ter correcting 50 p% pet ; * ry would One errors per 5 pages. How many pages Wo" of the first print of the book? s per page ; : Solution: Let X denotes the number ofan given by Then the mean numberof errors per PARE SS oOCdi Nea WIUI CallloUe144° = Probability, Statistics and Random Processes Using Poisson distribution, Probability that there are x errors per page = P(X =) OF 20,123, x! P(X =x) = P(x) The expected number of pages with x errors per page in a book of 1,000 pages is fe) = 1000x P(x) The recurrence relation for probabilities of Poisson distribution is a P(x+l) =~ P(x (+1) zl (x) P(0)=e7 wet = 0.6703 £(0)=1000 (0.6703) = 670.3 ~ 670 PQ)= Aro = (0.4)(0.6703) = 0.26812 £(1)=1000 (0.26812) = 268,12 = 268 P(2)= Aro = (026812) = 0.053624 £(2)=1000 (0.053624) = 53.624 ~54 P@)= 4 re) 0. a (0.053624) = 0,0071298 = 1000 (0.0071298) = 7.1298 =7 SC Pay= 4 PG) 0.4 4 (0.071298) OCANNEU WIL CalloceF(4) = 1000 (0.00071298) = 0.71928 ~1 The expected number of pages with x errors is obtained as No. of errors per page, x 0 i a er Expected number of pages, f(x) 670 268 54 7 1 EXAMPLE 4.18 fita Poisson distribution for the following data and calculate the expected frequencies: x 0 1 2 Bees f 109 65 2 3001 Solution: Here, mean of the Poisson distribution = Mean of the given distribution Lxf, Ane D7 N = Sf, =109465+22+3+1= 200 _ 04654444944 7 200 a ~ 200 A=061 Using the recurrence relation for probabilities, we can find P() using, pore Pe) The expected frequencies are obtained by, (fis) = PQ) The first probability is obtained using, P(0) =e =e" = 0.543351 | ‘f(0)=N-- P(0) = 200(0.543351) | = 108.6702 ~ 109 = 2 (0.54335) Scanned with CamScz146 © Probabitity, Statistics and Random Processes f(l)=N- P(l) = 200(0.331444) = 66.2888 ~ 66 Aa P(2) = — PU P(2) Isl (1) 7 OSt(o.331444) = 0.10109 FQ=N- PQ)= 200(0.10109) = 20.218 ~ 20 a = p2 PG) a = oS%.0.10109) = 0.20555 £@)=N -PG) = 200(0.020555) 11L~4 P(4)= 4 PQ) 61 = 81 0.020555) 4 = 0.003135 £(4)=N -P(4) = 200(0.003135) =0.627~1 Hence, the expected frequencies are: x 0 1 2 3 4 Total Observed frequencies, f 109 65 22 3 1 200 Expected frequencies, f(x) 109 66 20 4 1 200 Scanned with CamScz44 GEOMETRIC DISTRIBUTION Suppose we have a series of independent trials or repetitions and on each repetition the probability of success remains the same. Then the probability that there are x failures preceding the first success is q* p. Definition Arandom variable X is said to follow geometric distribution if it assumes only non-negative values and its probability mass function is given by q’p, x=0,1,2,3,..0) Scanned with CamSczMoments of Geometric Distribution = Drax PX =x) = Derg P = qr = pq{l+2q+3q° +--} = pal-ay" Hi = B(X?) = Dia PK =x) Hy = Lake -)+xi0"P = Diab -Di'p + erg P =237 a Y) je-tg2y 4 4 q apt 7PYD =2pgl-gy +4 P = 2prg? + A Variance of X is given by, V(X) = E(X*)-[E@)P 2 -£ Scanned with CamSczMoment Generating Function of Geometric Distribution EXAMPLE 3.82 Find the moment generating function of geome distribution and hence find its mean and variance. ; [AU May 706, June '07, April Solution We know that, M(t) = Ele*) = “ye p(x) = Leen Ewe ) p= rw r=0 a + get Ger? toe] =p(l - ge'y! = P (1—qe') Scanned with CamSczEXAMPLE 3.84 If the probability is 0.05 that a certain kind of measuring device will show excessive drift, what is the probability that the sixth of these measuring devices tested will be the first to show excessive drift? [AU December "07 Solution Given: p = 0.05, q = 1 - p= 0.95 and x = 6 Scanned with CamScza 226 © Probability and Random Processes oY We know that the probability: that the first success occurs ¢ om the PX = x)= 4" p, v=123 ath hit PX = 6) = (0.95)(0.05) = 00387 EXAMPLE 3.85 If the probability that the ¢: largct is destro shot is 0.5, what is the probability: that it will b oVed on any 1 destroyed on ny the 6th aten TAU June gy Solution The probability that it will be destroyed on sixth attempt is 4p (first five attempts are failures and sixth one is a success P= probability of hitting the target = 0,5 q=1-p=1-05=05 4°p = (0.5) = 0.0156 EXAMPLE 3.86 If the probability that an applicant for a drivers license Will pass road test on any given trial is 0.8, what is the probability that he wi finally pass the test . (1) on the fourth trial, (ii) fewer than 4 trials? Solution Let X denote the number of trials required to get the first success Then X follows geometric distribution, PX =x) ="! p, x= 1, 2,3, Given: p=08, .g=1-p=02 (i) The probability that the driver will finally pass the test on the fourth trial is (first three are failures and fourth test is a success) P(X = 4) = gp = (0.2) (0.8) = 0.0064 (ii) The probability that the driver will pass the test in fewer than 4 tls is P(X < 4) = P(pass in the first test or second test or third tes) (X =1) + P(X = 2) + P(X = 3) =p+qp+ ap -992 = pl +4 +42) = (0.8) [1 +.0.2 + (0.2)°) = 0% we at is EXAMPLE 3.87 A dic is thrown repeatedly until 6 appears. Wh probability that it must be thrown more than 5 times? Solution p = probability of success , ; = probability of getting number 6 in throwing a die a Pr“ 6 5 ~pal-555 and ocalineu will GamSceStandard Distributions > 227 The probability that the di ite should be thrown more than 5 1 S times (first fiv e are failure) =v _ a + gp + q?p + = gp +g tgs.) = gpl -q)! = q pp) = g 3\ = (3) = 0.4019 Aliter The probability of getting 6 i and q Let X denote the number of thr rows fc By geometric distribution ee PX =x) = Gop, x=12,3,. Since 6 can be got cither in first, second, ... throws P(X > 5)=1- PUSS) QTC) Ley = 0.4019 6 " ain until the first head occurs ‘Assuming the ability of a head occurring 1s P. find the id number of tosses 1S required is probability is 0.5 that an EXAMPLE 3.88 A coin is tossed tosses are independent and the pro! value of p so that the probability that an equal to 0.6, Can you find a Value of 7 $0 that the odd number of tosses 1S required” Solution Let X denote the number of t metric distribution (success). Then ¥ follows 8¢° josses required to get the first head given by OCAMNEU WIL UalloCce228 ¢ Probability and Random Processes PIX = x)= pg! x =1,2,3,.., P(X = an odd number) = POX = 1 or X= 3 or X= 5, ) =PW=1) + P= 3) + PRS) y =p + pq + pat + page + -.. - =p +g + gh + gh + = pl-4’) ee (=) (+4) ig’ Pao Given: P(X = an odd number) = 0,6 = —— l+q = 6 (1 +g) = 06 + 0.64 > -06=04 0.4 2 3 21 and = 1l-g=1-=== Peet 33 Next, if P(X = an odd number) = 0.5 = —— 05(1+g)=1 > 059-05 q=l, p=l-q=0 Note: When P(X = x) = constant, the discrete random variable X is said follow a discrete uniform distribution. EXAMPLE 3.89 _If the probability of success on each trial is 0.25, after how many trials can we expect first success? ~ = Solution Given: p = 0.25,q =1-p=0.75 IX denotes the number of trials required to get first success, then by geometnt distribution, expected number of trials for first success is E(X)=4=423 P Blea le «beats? EXAMPLE 3.90 Let one copy of a magazine out of or oa special prize following geometric random distribution. Det av ‘june ‘tl variance. Sle Solution Given: ei ocalineu witli CamSce2Standard Distributions © 229 G=1-p=y-1_9 10-10 Mean of the geometric distribution is 4 — 9 P Variance = F192 _ gg P10 EXAMPLE 3.91 Suppose that a trainee soldier independent fashion, If the shoots a target in an Probability that the target is 0.8 is shot on any one shot () What is the probability that the target would be hit on 6th attempt? (ii) What is the Probability that it takes him less than 5 shots? (iti) What is the probability that it takes him an even number of shots? Gx) What is the average number of shots needed to hit on target? [AU December *05] Solution Given: p=08 and q=1-p=1-08=02 If X is a random variable denoting the number of shots required for the first success, then using geometric distribution PIX =x)=G''p, x= 1,23. > P(X = x)=(0.2"-! 8), x=1,2,3,.. (i) P(the target would be hit on the 6th attempt) = P(X = 6) = (0.2)6 # (0.8) = (0.2)* (0.8) = 0.000256 (ii) P(it takes him less 5 shots) = ma= 1 =3) + P= 4) = =2) + P&=3) Ra Doe BE Bg afl + (0.2) + 027 + 02) = 0.9984 cots) (iii) P(it takes him an even number of sI “s yee = P(X = 2) + P= 4) + PC = 9# : 0.2)? 10.8) + 0 24-10. i fe) os) + 0.2 + oats hott + (0.04) OOF ay = piel 004)! -9 . . = oe an hots necded to hit the target (iv) Average (expected) number of s! 0. i} OCadlineu wiuTCaMTScewer 230 ¢ Probability and Random Processes EXAMPLE 3.92. A and B shoot independently until ©2ch has his oyn The probability oftheir hitting the target at cach shot is 3/5 and 5/7 regye {ange Find the probability that B will require more shots than A. clive Solution Suppose A requires X number of trials to get his first Sucee 85, . 3 ee a with p=2.q=1-5 =<. then X follows geometric distribution given 1 _(3)(29 P(X =r) = Pq -(3)(2) r=1,2,3,.. Suppose B requires Y number of trials to get his first success, p= 5, qai-3 ; then Y follows geometric distribution given by rd PY =N= pag! -(3)(2) r=1,2,3,.. 7 The probability that B requires more trials to get his first success than 4 requires to get his first success is that if A requires r trials to get his first success that B requires r + 1, 7 + 2, ... trials to get his first success. = DP =r and Yar thr +2002) = SPX =r) PU =r + Dor (Y= + 2)..] by independence Oe ocalineu will GamSce2EXAMPLE 3.94 If the MGF of X is (5 — de’), find the distribution of ¥ X= § or 6) [AU December *04] Solut Pua aoa sane anne “ton The MGF of Geometric distribution is “1 (He) = (5 — 4e") | (given) p Scanned with CamSczU and p= Ala ale q= The PMF of geometric distribution is PIX =x)=G¢° !p, Ose aaa P(X = 5 or 6) = P(X = 5) + P(X = 6) =gp +a p=pgil+9 36) = 0.147456 Scanned with CamScz4.6 UNIFORM DISTRIBUTION This is a discrete distribution and is defined as follows: Definition A random variable X is said to have a di: defined by Ss probability mass function crete uniform distribution if adh Scanned with CamSczStandard Discrete Distributions = 153 1 PU ayy a} PN 4 (4.15) 0, otherwise Caution: «© Thi ‘one where the random Variable assumes. + Here, is the parameter of the distribution, n takes only set of positive integers, When the different values of the random + The distribution is appropriate for a di each of its values with equal probability nm Variable become equally likely, the distribution can be used. ie experiment and an experiment with a deck of cards. Moments of Uniform Distribution Mean of uniform distribution is given by, EX) = Ei xP(X =x) -=a(! E(X)= (Z)ztox (383) n i E(X?) = Dix P(X =x) = dia 1 ninsnneD) “Ge (n+ (2n+1) a The variance of uniform distribution is given by, ; V(x) = BU) -[BOO - orton (221) a Coenaned ey 2(n+N(2n+)- 30+" re Scanned with CamScz154 Probability, Statistics and Random Processes _ (n+1)(4n+2-3n-3) ~ 12 (nt )(n-1) VX)= The MGF of Uniform Distribution Let \’be a discrete random variable, which follows uniform distribution. Then the MGF of Xis givens M,(t)= Ele") = Lie" PX =x) » ox(l -r.e(!) n Ln oe =p de a 7 ete] 1 = Fle +e el) +... (et)"] I = Tere tev tice! yr] Scanned with CamScz