Probability nn 5.1 Some Fundamental Concepts ‘Sample Space and Event: Consider an experiment whose outc an experiment is called a random experiment. However, although ‘known in advance, let us suppose that the set of all possible outcomes of an experiment is known as the sample space of experiment an ‘Some examples follows: ome is not predictable with certainty. Such the outcome of the experiment will not be ss known. This set ofall possible outcomes \dis denoted by S. (} the outcome of an experiment consist inthe determination ofthe sex o! a newborn chid, then ‘S={g, b} where the outcome gmeans thatthe child isa gi and bis the boy. (i Hithe outcome of an experiment consist of what comes up ona single dice then = (1,2, 34,5, 6} (i) I the outcome of an experiment i the order of irish ina race among he 7 horses having post positions 1,2, 3, 45,6, Z;then S = [all7! permutations ofthe (1 2.3, 4, 6, 7)} The outcome (2.3, 1.6.5, 4, 7) means, for instances, thatthe number 2 horse comes in fist then the number 3 horse, then the number horse, and soon. ‘Any subset E of the sample space is know as Event. Thats, an eventis a set consisting of some or all af the possible outcomes ofthe experiment. atthe outoorne ofthe experiment is contained in E, then We say tha E has occured, Aways ES. Sinoe € and'S are sets, theorems of settheory may be effectively used to represent and solve probability reve complato. ; os Fone posts te preceding exale()\"E,= (gl, hen& isthe event that he cls a gi =(0) ne Fe ethan one outcome. Such as E =|, 3,5] for an experientf towing a cce Wo sy event may consist of mot See ifthe dice comes up 1 or3 or. Ehas happened ithe clot Pana F ola Bama apace Swe dele the new event E UF to consis of ll For any te per in E orin For in both Eand F, Thats, the event E UE wil occur it ether E or For both outcomes that or on ice example (Dif event E,= (1,2) and E = (9,4), then EUF = (1,2, 8,4). occurs. For instances.Thatis E UF would be another event consisting of 1 or2or3 or 4, The event EUF iscaled Union of ven, Eand the event Similar forany two events E and F we may also done the new event EAE Caledintersectn of E and F to consists ofall outcomes that are common to both E and F s $ 5 F WP DOI (©) Sradesregon:EWF (@)Shadedregin:ENF —(@)Shadedregonce™ ECF ‘Mutually Exclusive Events ‘Two events E and F are mutually exclusive, HEAF = gi.e.P (EnF)=0. nother words, cannot occur and if F occurs, then E cannot occur (ie. both cannot ‘occur together). Collectively Exhaustive Events it Boccurs, ¢ outcomes, P(E UF) = P(S) = 1 DeMorgan’s Law nF oa @ (Se) = her O) (as) = Ler Example:(E,W E,)° = EF AES (E, VEL = EF UES Note that EF n EF is called neither E, nor &y E, uF, is called either E, or E, (or both). 5.1.1 Approaches to Probability ‘Thete are 2 approaches to quantilying probabilty of an Event E. 1. Classical Approach: P(E) = 7. if 5.1.2. Axioms of Probability Consider an experiment whose sample space i S.For each event E Of the sample space S we assume that a number P(E) is defined and satisfies the following three axioms,pone: 08 PCS? axiom-2: ps)=1 ‘pgion-3: For any sequence of mutually exclusive events E,, E>,.....that is, events for which , 9 E)=@ ten] cp us) 5PE) mat) iat example: RE, U E,) = PLE;) + FLE,) (Ey E, are mutually exclusive) gon siete Propositions tristo be noted that E and E® are always mutually exclusive and since EVE” = S, We have by Axiom (a) that : PIE E* (E) + P(E) = P(S) = 1 roposition-1: P(E® Proposition-2: If E (159 + 160) = 169.5, yonfor Grouped Date 4. Identity the median class which contains the middle observation (=:(n-+1/2)"ebservation). This. can be ore by observing the first class in which the cumulation frequency is equal fo or moro than ne 2, Calculate Median as follows: ‘mn Where, Lower limit of median class Total number of data Items = ZF = Cumulative frequency of the class immediately preceding the median class fy# Frequency of median class h=width of median class Median for Grouped Data ‘Consider the fllowing table giving the marks obtained by students nan exam a 7 [oamaaive MarkRange | no, ofStudents | Frequency. 0-2 2 2 20-40 3 A 40-60 10 6 60-60 15 30 90-100 2 50 N41 NET 55 Here st Te dana o-oteierredancinsancecun-on sl? 255 ooe[2. Median = : spesan mua tne css senso 607 (0) 6 (atlees apts le han 607 ad (tos al gamer tan eB mas ie. (atleast4 Mode and Standard Deviation ‘Mode: Mode is defined as the value of the variable which occurs most frequently. Calculation of Mode: Mode is that value of x for which the frequency is maximum. if the values ‘grouped into the classes (such that they are uniformly distributed within any class) and wo have a frequen distribution then cy (Identity the class which has the largest frequency (modal class) (i) Calculate the mode as: Ores Mode = L+— 971, 2 Where, Lower limit of the modal class Largest frequency (frequency of Modal Class) f,= Frequency in the class preceding the modal class Frequency in the class next to the modal class. Width of the modal class Data relating to ight of 352 school students are given in the following | frequency distribution. Calculate the modal height. | Heigh [Number of (infeet) | students 30-3: 2 35-40] 397 40-45] 79 45-50] 152 50-55} 65 55-60 7 | Total 352 Solution: Since 152 is the largest frequency, the modal class is (4.55.0). Thus L = 45, f= 152, f, = 79, f, = 65,h= 05. 152-79 Mode = 4.5 + = x05= i lode = 45 + eee gg X05 = 473 (approx) While mean, median and mode are measures of central tendency. 5.5 Standard Devi Standard Deviation is a measure of dispersion or variation amongst data. Instead of taking absolute deviation from the arithmetic mean, we may square each deviation and 0b!” the arithmetic mean of squared deviations. This gives us the ‘variance’ of the values, ‘The positive square root of the variance is called the ‘Standard Deviation’ of the given values-~ peviation from Raw Data sot ‘supposes =a" #9 E7 values of the x, their arithmetic mean ig; 1 HF, Fly -¥ ‘ = fiaand 1-Fet2 Fan are the deviations of he values otto. Then a au at is the variance of x, It can be shown that wa _ 15? 52 _ MBs (Ea 2 Ba = oa} aoe a itis conventional to represents the variance by the symbol 2, infact, ois small sigma and Z's capital a. S97 yare oot ofthe variance isthe standard deviation gen ats — =P = [fxe-# - jut bap ‘aladation of Standard Deviation from Grouped Data, Calculation for standard deviation for grouped data can be shown by this example: ‘The frequency distribution for heights of 150 young ladies in a beauty contests given below for wich we have to caloulate standard deviation. Height | Mid values | Frequency Gxxy x (ininches) | x ' xa | hot ‘47250.75 620-635| 62.75 12 753.00 | po561.05 635-65.0| 64,25 20 1286.00 | 49 1045,75 65.0-66.5 65.75 28 1841.00 81406.125 66.5- 68.0 67.25 18 1210.50 29806.125 68.0- 69.5 68.75 19 1306.25 89804.6875 695-71.0| 70.25 20 | 1405.0 | osro 26 T10-725| 71.75 30 a 154441.875 725-740| 73.25 3 2 16096.6875 Total 150 10173.00 | 691308.375, yg = EhH 2 10173 ~ 67.82 Thus, oy et Blah 2 46087205 and zh Nez = 100 Therefore, the variance ofS : og = Dish gt MEME Bla 5704 «9, = 8.03 (Inches) and standard deviation is 3.03 inches).5.6 Random Variables Itis frequently the case when an experiment is performed that we are mainly interested in. 8078 tre, of the outcome as opposed to the actual outcome itse. For instances, in tossing dice we are often interested in he sum of two dice and are not ratty con, about the separate value of each die. That is, we may be interested in knowing that the sum is 7 ang Pot be concerned over whether the actual outcome was (1.6) or (2, 5) oF (3, 4) or (4, 3) or (5, 2) or (6, 1) : Ao, n coin fpping we may be interested in the total numberof heads that OCCU ard not care ata about the actualhead tl sequence that results, These quantities of interest, or mor formal, hese rel vay functions defined on the sample space, are known as random variables, Because the value random variables determined by the outcome ofthe experiment Wemay as probabilities to the possible values of the random variable. ‘Types of random variable: Random variable may be discrete or continuous. Discrete random variable: A variable that can take one value from a discrete set of values. Example:Letx denotes sum of 2 dice, Now xis aciscrete random variable asitcantako one value tom, the set {2, 3, 4, 5,6, 7, 8,9, 10, 11, 12], since the sum of 2 dice can only be one of these values, Continuous random variable: A variable that can take one value form @ continuous range of value, xample:x denotes the volume of Pepsin a 600 mi cup. Nowxmay be a number from 010500, any which value, x may take, 5.7 Distributions Based on this we can divide distributions also into discrete distribution (based on a disc random variable) or continuous distribution (based on a continuous random variable). Examples of discrete distribution are binomial, poisson and hypergeometric distributions. Examples of continuous distribution are uniform, normal and exponential distribution. Properties of Discrete Distribution © SPlx)=4 + Ele) =20 Fa) + Vi) = Ete) (Eta)? * Vos) = 322A) [Ex Pe)? Properties of Continuous Distribution © freee a1 F(a)= J f(x)d (cumulative distribution function) Ee)= Ff xt(a)ar _ a 2 Va)= f Pr-| J oa » Placx 0, ott eed For Poisson distribution: Mean = Et Variance = Vi Therefore, expected value and variance of a Poisson random variable are both equal o its parameter), EEE | 4 certain airport receives on an average of 4 aircrafts por hour. What a the probability that no aircraft lands in a particular 2 hr period? Solution: a= rate of occurrence of event = 4/hr = average no of occurrences of event in specified observation period At= aAt Inthis case a= 4/hr and At= 2h Bs Ae 4x t=8 Now we wish that no aircraft should land for 2 hrs. Le. x= 0 850 oO Frequently, poisson distribution is used to approximate binominal distribution when nis very large and p is very small. Notice that direct computation of nC, p¥ (1 - p)?-* may be erroneous or impossible when nis very large and pis very small. Hence, we resort toa poisson approximation with = np. TEES 4 certain company sells tractors which fail at a rate of 1 out of 1000. 500 tractors are purchased from this company what is the probability of 2 of them failing within Solution: = m0 500% Yoon = Js aye riea2)= UE) son 5.72 Types of Continuous Distributions 1. Uniform distribution 2. Exponential distribution 3. Normal distribution 4, Standard normal distribution 1. Uniform Distribution |n general we say that X is a uniform random variable on the interval (a, B) it its probability density function is given by:: te ——__— fay= [Fra Hasxep 0 othenwise singe #6) sa constant, ll values ofx between «and are equal tkaly (fon Graphical representation: . fe) pls —, Ba a be 8 For discrete uniform distribution: Mean = efx) = #2 2 i (6-«)? Variance = V(X aiance = Vix) = OS FX is uniformly distributed over (0, 10), calovlate the probability (a) X<3 (b) X>6 (©) 3 O by A continuous randor iable whose prob: ie itx20 f=) — itx<0 sit random variable with parameter. The cumulative cstibutve function F(a) of i nen is said to be ‘expor ones aven by: ym vari _ an exponential rand [tretdc= (o™]f =1-em azo‘Suppose that the length of a phone call in minutes Is an exponential random variable with parameter A = 1/10. If someone arrives immediately ahead of you at a public telephone booth, find the probability that you will have to walt, (a) More than 10 minutes (b) Betwoon 10 and 20 minutes Solution: Letting X denote the length of the call made by the person in the booth, we have that the desired probabilities are: @ P(X> 10} = 1-P(< 10) = 1-F (10) = 1- (1-649) eh =o1=0.968 (b) P(10 otherwise ilacx

atleast one of A and B ocours is 3 (a) 04 (b) 0.6 (©) as do 1 (@) 088sity function of a variable “ opabily dr os ™P x 4.2.3-4.5.6 POX) K 3K 5K 7K 9K 11K 13K a to vau0 - 1 Os Op 50 i 1 ae On ® or po 1 1 OG Om 9 Ina lottery, 2 tickets are drawn at a time out of6 tickets numbered from 1 06. The expected value of the sum of the numbers on the tickets drawnis @7 (o) 6 ©@s @4 Q.10 Two dice are thrown simultanoously. The probability that atleast one of ther wil have facing up is al al alr au @ % Os 25 at 3 Os eit iat bs w conto eres EE with folowing distribution osxs2 ffx) = ke Z are respectively The value of kand PUT x52) Lf Oa ) h i 12 Op5 @ Ht Q.12 A gambler has 4 colns in hor pockot. Two aro doublo-headad, ona is doublo-talad, and ona is normal, Tho coins can not bo distinguished ‘unless one looks at them, Tho gamblor takes a ‘coin at random, opens her eyes and soes that tho uppor face of the colin is a hoad. Whats tho probability that the lower face Is a head? 8 4 ) 5 Os 2 4 Og 4 Q.13 Let X be uniformly distributed on (0, 1, 82) whats Pr[3x + 12 0(mod 33)]? 1 @ 3a ) Z og ot 1.14 Suppose you are given a bag containing 9 unbaised coins you are told that n— 1of these are normal coins, with heads on one ‘side and talls on the other; however the remaining coin has heads on both its sides. ‘Suppose you reach intothe bag, pickouta coin uniformly at random, flip itand get a’ head. What js the (conditional) probability that this coin you chooses the fake (|.0., double-headed) coin? 1 2 on od oz o% rs {Q.48 Two randomvarlables Xand Yare independent ifthe pa of events X and Yjare independent no rmattar how you choose the values ! and j, Which cof the following most accurately expresses the propostion that Xand Yare nat independent?(6) for all, j ProLX and ¥) # PrIX) PrlY}) (©) forall, somo j PrfX,AND ¥,]#Pr 1X) PY) (©) forsome alli PX AND ¥] # PrIX] PAY) (0) for some, j Pr[X AND ¥,] # PrlX] PLY] Q.16 Suppose you are given a bag containing n unbiased coins. You are total that n- 1 of these are normal coins, with heads on one side and tails on the other; however, the remaining coin thas heads on both its sides, ‘Suppose you reach into the bag, pick outa coin uniformly at random, ‘Suppose you flip the coin ktimes after picking it (instead of just once) and see Kheads. What i ow the conditional probabil that you picked the fake coin? ie zt ® ae © Ta © = 7 (6) None ofthese (Q.17 The expectation and variance of@ random variable X, +X where X, and X, are independent random variables with expectation and variance o. @ no (0) w20 (©) 2,08 (@) 2u,20 .18 Foreach square of an 8x8 checker board, fpa fair coin, and color that square black or red according to whether you get heads or tails, ‘Assume that all coin fips are independent. A same-color row in a row on the board were all ‘squares in the row have the same color (Le, all ‘ed, oF all black). Let the random variable X kao S252) = fPM(a)dx 24 3 oe a (b) There are 5 faces that are heads out of a total of 8, so the probability is 3. Let A be the event that the upper face isa head, and B be the event that the lower face is heads. PrfA] = Pr{B] = 3 PAN B] = é = 3 el So, PrBIA] = a z.4 8 (d) 3X +12 = 0(mod 33) * 8X = ~12 (mod 33)3X = 21 (mod 33) X=7mod 11 2 2 2 yen 7, iBand.29 tow, 4 1220(mod 33)] 2 prllr=7) or (e= 18) or (x= 29)] zptic= 7] + ple = 18] + pfx = 29} 14, 0) Thetvee diagram with probabilties forthe given problems shown below: 1 hove ae F couble headed coin _1_ Heads dna p{double head coin|Heads) = ——2__ x Ta aru > " fal > z 16. (c) ' ‘The tree diagram with probabilities for this problem is shown below: 4 oS Normalcoin BE knead cate Carne «ple headed coin | kheads out of Kirials) 1 x! at = pet Ty iy an 20. MX, and X, are independer Putting, iia) X)+ VX) =O 4or=2gt . (a) Pa tow being same ‘color) nae ow being all black) + p(a row being ail “=the JG pr(X = 0) = pr(0 out of 8 rows being of same colour) ). (b) ‘The tree diagram with probabilities is 4 un coin > Z_ kowotkheads 2 headed coin outofk heads ‘HXprocedureisin enor) = lnommalecinand kout of kheads) + p{fake coin and not koutof kheads) oo (3) +tx0 1)_(1¥ @-1) “ka) a (a) ‘The artival pattern follows poisson dstibuton. piK=x)= yo Here heat where, «= number of events/unit time = 20fhr 3 At= 45min = 5 min = hedeaat=20x 2 =15 a 21. (b) we [lpr “fee pe SL -w) Peles = afenye _ | Otrisecd ae =) triseven a Ura «. odd central moments are zero, even central ‘moments are etapetan=4 Wee Ais Bebe Fe 22. (d) Here, w= Oando= B Using Bienayme-Chebyshev rule, ko SxS + ko)21— Ph H+ ko)2 1-25 1 Pho SxS + ko)21- ko = ke . 6 - i> 23. 24. 25. (b) The tree diagram is showm below: o/ Netinda 201 tnrgsrnin 035, o South India {roe registration ke Wesem nga 222 tes regstaten pitree registration) = 0.15 0.01 +0.35%0.05 + 050 x 0,02 =0.029 (a) ASouth India free registration) _ p(South india and free registration) = ‘P(free registration) = osx = 0.6034 ~ 60% (b) p=001, n= 100 Using poisson approximation to binomial distribution, 2.=np=100x 0.0