this chapter, you should Wena cera Seaeeamne cente Sencha tes aot provi and asta probability, aloes, taurine Understand the addition theorem, the ‘rultiptcation theorem and other important ‘on probability , and their use in ims in various diversified Tree th neof pray ther Ms tie ean a Sach reece. CHAPTER OUTLINE | 31, INTRODUCTION 3:2. SHORT HISTORY 33. BASIC TERMINOLOGY 34. MATHEMATICAL (OR CLASSICAL OR ‘A PRIORI’) PROBABILITY 3-4-1. Limitations of Mathematical Probability 35. STATISTICAL(OR EMPIRICAL) PROBABILITY 35:1. Limitations of Empirical Probability 36, SUBJECTIVE PROBABILITY , 3-7. MATHEMATICAL ‘TOOLS: PRELIMINARY NOTIONS OF SETS: 37-1. Sets and Elements of Sets 372, Operations on Sets 37:3. Algebra of Sets { 374. Limit of Sequence of Sets 375. Classes of Sets. 38. AXIOMATIC APPROACH TO PROBABILITY 3.8-1, Random Experiment, Sample Space and Elementary Events 38-2. Event 3.83. Acceptable Assignment of Probabilities q 38-4, Natural Assignment of Probabilities Mn1718. ra contributors are: T. Bayes (Inverse probability, Pre Laplace (1749-1827) ire esearch over a number of years finally pebliahca “Theorie Russian mathematician also have made very valuable contributions to the modern, theory of probability. Chief conteibutors, to mennes only a few of them are : Crag chev (1821-94) who founded the Russian Seleor cy Statisticians; A, Markoff Nenu, 1222); Liapounoff (Central Limit Theorem) yA Khintchine (Law of Large Numbers) and A. Kolmogorov, who axiomised the cactus probability. 3:3. BASIC TERMINOLOGY rept to8 We shall explain the various terms which are used in the definition of probability under different approaches sae antom Experiment. ifn ench tril ofan experiment conducted under identical Senditions the outcome isnot unique, but may be any one of the possible outcomes, they sel ‘am experiment i called a random experiment, {Framples of random experiments are: tossing a coin, throwing a die, solecting card from a pack of playing cards, selectin a given group of families, te: In all these cases, there are a number o which can oceur but the isan uncertainty as to which one of them ater (0A deisa small cube used in gaming, Onis si oes, dots ae marked as Marl of eee The tome of throwing dei the nue of date on ie upper face Jos ae ant Consist of fursuit called Spies, hrs, Dm sna Chis. Bach si jack or ee ras otek nine cards are numbered rom See ing, a queen and Taeed cn"? SPades and clubs ate back faced corde hie Ren diamonds are faced cards. Gy Cateome. The result of «random experiment will be called an outcome, ‘abril! and Event. Any particular performance of random experiment is called Imeleed outcome ox combination of outcomes are termed nena For example, () Tf coin is tossed repeatedly, the result is not unique. We may get any of t two or fala s Os8ng OF a coin is a randam experiment or Meant ‘getting of a head or tail is an event faced die and observin, ‘henmber of points that appear, the posible outcomes 12,3,4,5,6 Tenthe same experiment, the possible events could alse he stated as Oil nie of pnts Ecen suumberof points’ “Getting a point greater than; and s0 on. dependent on iTHEORY OF PROBABILITY 35 Event is called simple if it corresponds to a single possible outcome of the experiment ‘otherwise itis known as a compound or composite event. Thus in tossing ofa single die the event ‘getting 6" sa simple event but the event of getting an even number isa composite event 4. Exhaustive Events or Cases. The total number of possible outcomes of random experiment is known as the exhaustive events or cases. For example, (i) In tossing of a coin, there are two exhaustive cases, viz., head and tail (the possibility of the coin standing on an edge being ignored). (ii) In throwing of a die, there are 6 exhaustive cases since any one of the 6 faces 1, 2, 6 may come uppermost (ii) In drawing two cards from a pack of cards, the exhaustive number of cases is %2C,, since 2 cards can be drawn out of 52 cards in 1C, ways. (io) In throwing of two dice, the exhaustive ntimber of cases is 6 = 36, since any of the numbers 1 to 6 on the first die can be associated with any of the 6 numbers on the other die. In general, in throwing of n clice, the exhaustive number of cases is 6" <3. Favourable Events or Cases. The number of cases favourable to an event in a trial is the number of outcomes which entail the happening of the event, For example, GV in drawing a card from a pack of cards the number of cases favourable to drawing of an ace is 4, for drawing a spade is 13 and for drawing a red card is 26. Gyn throwing of two dice, the number of cases favourable to getting, the sum 5 is (1.4), 4D, 2.3), 8,2), 12, 4. 6. Mutually Exclusive Events. Events are said to be mutually exclusive or incompatible if the happening of any one of them precludes the happening of all the others, ce, if no two or more of them can happen simultaneously in the same trial. or example, (i In throwing a die all the 6 faces numbered 1 to 6 are mutually exclusive since if any one of these faces comes, the possibility of others, in the same tral, is raled out. (i Simitarty in tossing a coin the events head and tail are mutually exclusive. 7. Equally Likely Events. Outcomes of trial are said to be equally likely if taking into consideration all the relevant evidences, there is no reason to expect one in preference to the others, For example, (i) Ina random toss of an unbiased or uniform coin, head and tail are equally likely events. (ii) In throwing an unbiased die, all the six faces are equally likely to come. +8, Independent Events. Several events are said to be independent if the happening (or non-happening) of an event is not affected by the supplementary knowledge concerning the occurrence of any number of the remaining events. For example, 4’) In tossing an unbiased coin, the event of getting a head in the first toss is independent of getting a head in the second, third and subsequent throws. ~{ii) When a die is thrown twice, the result of the first throw does not affect the result of the second throw. If we draw a card from pack of well-shuffled cards and replace. t before drawing the second card, the result of the second draw is independent of the first draw, But, however, if the first card drawn is not replaced then the second draw is cependent on the first draw.36 FUNDAMENTALS OF MATHEMATICAL STATIST 3.4, MATHEMATICAL (OR CLASSICAL OR ‘A PRIORI) PROBABILITY Detinition. If « random experiment or a trial results in ‘n exhaustive, mut exclusive and equally likely outcomes (or cases), out of witch m ave favourable to ‘occurrence of an event E, then the probability ‘p” of accurrence (or happening.) of E, ust denoted by PCE), is given by = Number of frvourable cases p=PE) = real neal yaa aos ‘This definition was given by James Bernoulli who was the first person to obt quantitative measure of uncertainty. Remarks 1. Since m) 20, >and m <0, we pet from 3-1) P(E)20 and P(E)St = OSPIE)ST 2, Sometimes we express (31) by saying that the odds in favour of Fare m : (~m) 6 ‘odds against E are (1 ~m) =m, The non-happening ofthe event E is called the complementary exent of E and isd by Eor £& The number of case favourable to E , ie, non-happening of Eis (=m) probability that £ will not happen is given by q= P(E) = 7 > pte G-1-PQ > Pw=1-rB or MH +E Ueihe event repress the happening of atleast one of the events Ey Ey Egil complementaty event E represents the happening of none of the events Ey, Ey a Substituting in (31), we net A P (Happening of atleast one of the events E,, Ea,» E,) =1-P (Noneo the events Ey EE, happens) 4. Probability’ ofthe happening of an events aso known as the probability of sue the probability “7 of the non-happening of the event as the probability of failure, {2 p+ 5.1f P(E) = 1, Eiscalled a certain event and i P(E) = 0, Eis called an impossible event. 6. We can compute the probability in (1) by logical reasoning, without conduct ‘experiment. Since, the probability in (3-1) can be computed prior to obtaining any experim data, itis also termed 38's prion” or mathematica! probity __3-4-1, Limitations of Classical Definition. This definition of classical probab breaks down in the following cases + (j) Ifthe various autcomes of the random experiment are not equally ‘equally probable. For example, (a) The probability that a candidate will pass in a certain test is not 50%, sit two possible outcomes, viz., success and failure (excluding the possibility compartment) are not equally likely. (0) The probability that a ceiling fan in a room will fall is not 1/2, since the of the fan ‘falling’ and ‘not falling’ though mutually exclusive and exhaustive, ‘equally likely, In fact, the probability of the fan falling will be almost zero. (0) Ifa person jumps from a running train, then the probability of his survival ‘not be 50%, since in this case the events survival and death, though exhaustive mutually exclusive, are not equally likely.THEORY OF PROBABILITY 37 (ii) If the exhaustive number of outcomes of the random experiment is infinite or unknown. 3.5, STATISTICAL (OR EMPIRICAL) PROBABILITY Definition. (VON MISES). If an experiment is performed repeatedly under essentially Iomiogeneous and identical conditions, then the limiting value of the ratio of the muniber of times the event occurs to the number of trials, as the nuniber of trials becomes indefinitely | large, is called the probability of happening of the event, it being assumed that the lint finite and unique. ‘symbolically, if in N trials an event E happens M times, then the probability of the happening of E, denoted by P(E), is given by : PE) = him + (2) Remarks 1. Since in the relative frequency approach, the probability is obtained objacteel of tpettine empirical ebosvaltons: Ws aarEnvwn a Expr Pro. 2, An experiment is unique and non-repeating only in the case of subjective probability. In other cases, there are a large niumber of experiments or trials to establish the chance of dccurrence of an event. This is particularly so in case of empirical probability. In elassiea! probability also, repeated experiments may be made to verify whether a deduction on the basis {iy certain axioms oF undisputed lnwys is justified. Only after repeated trials it can be established Sgt the chance of head in a toss of a coin is 1/2. J, E. Kerrich conducted coin tossing, experiment with 10 sets of 1,000 tosses each during his confinement in World War Il, The number of heads found by him were: 502, S11, 497, 52%, 504, 476, 507, 520, 504, 529. This gives the probability of getting a head in a tuss of a coin ax: S07, = 05079 = 3 Thus, the empirical probability approaches the casical probability asthe number of tal large. i finitely 5-1. Limitations of Empirical Probability, (i) If an experiment is repeated a large number of times, the experimental conditions may not remain identical and homogeneous. (i) The limit in (32) may not attain a unique value, however large N may be. pare 8-1. What is the chance that a Teap year selected at random will contain $3 subays ? Solution: tn a leap year (which consists of 366 days), there are 52 complete weeks and 2 days over. The following are the possible combinations for these two ‘over’ days : (@) Sunday and Monday, (ii) Monday and Tuesday, (iif) Tuesday and Wednesday, (io) Wednesday and Thursday, (0) Thuftsday and Friday, (vi) Friday and Saturday, and (cil) Saturday and Sunday. In order that a leap year selected at random should contain 53 Sundays, one of the two ‘over’ days must be Sunday. Since out of the above 7 possibilities, 2, viz, () and (ii), are favourable to this event. Required probability =3 etn “Two unbiased dice are thrown, Final the probnbility that: (i) both the dice show the same number, (ii) the frst die shows 6, Ass FUNDAMENTALS OF MATHEMATICAL STATISTI (ii) the total ofthe numbers om te dice is 8. (ie) the total of the numbers on the dice is greater than 8, (2) ‘the total of the numbers on the dice is 13, and (i) the total ofthe numbers on the dice is any number from 2 to 12, both inclusive. Solution. Ina random throw of two dice, since each of the six faces of one die be associated with each of six faces of the other die, the total number of cases i 6X6 =36, as given below : an @n (2) 2) “3 @3) 4) Qa) (5) 25) G5) 45) 65) a) 26) 66 6) 66) Here, the expression, say, (i ) means that the first die shows the number i and Second die shows the number j. Obviously, (i,j) # (if #) s Exhaustive number of cases (n) = 36, (a) The favourable cases that both the dice show the same number are (1,1), @ 23,3), 4,4), 6, 5)and (6,6), ie, m =6 <, Probability that the two dice show the same number = £ (8) The favourable cases that the first die shows 6 are (6,1), (6,2),(6, 3), (6.4), (6,5) and (6,6), ie, 6in all. t “+ Probability that the first die shows ‘6’ = =1 =£=1 (c) The cases favourable to getting a total of 8 on the two dice 2,6),(3.5), (4,4), (6,3), (6,2), dencm=5, Probability tha total of numbers on two dice is 8 =$, en (3,2) G3) Ga) (4.1) (4,2) (4,3) (44) (51) (5,2) 6.3) (54) 1) 6,2) ,3) 4) (6,5) (6.6) 1 6 are : (1) The cases favourable to getting a total of more than 8 are: (3,6); 6,3). (4.5), (5.4), (4,6), (6,4), 6,5), (5,6), (6,5), 6,6), ie, m=10. Probability that the total of numbers on twa dice is greater than 8 = ¥= § 4 (This isan example of an impossible event, since the maximum total can be 6+ 6. 12, Therefore, the required probability is 0. (f) The probability is 1, asthe total of the numbers on the two from 2 to 12. The given event is called a certain ews. © \ Example 3-3. (0) Among the digits 1, 2, 3, 4,5 at first ome is clusen and then a se 4 selection is made among the reniining four digits. Assuring thot all teenty poss ‘cuctcomes lave equal probabilities, find the probability that an odd digit will be selected (0) the frst time, (ii the second time, and Git both times. (b) From 25 tickets, marked with dice certainly ran, first 25 mumerals, one is drawn ab random. Find Hl chance that (i) tis multiple of 5 or 7, and (iit bs a nuultipte uf 3 or 7. Solution. (a) Total number of cases = 54 = 20, THEORY OF FI (i) Now there are 14), (1,5), 16.20 (ii) Also there a 60,4,1,6.0,4 (0) @ Numbers Wand 25, ie., Sir inall. Hence 15,1821 La in all. Since the n distinct favourable, kking can beTHEORY OF PROBABILITY 39 (i) Now there are 12 cases in which the first digit drawn is odd, viz, (1,2), (1, 3), 4,15), 10,6, 2, 6 4) 8-9).6, (6,2) 6 3) and 6, 3 The probability thatthe first digit drawn is odd =22 =3 (ii) Also there are 12 eases in which the second digit drawn is odd, viz., 2, 1), (3,1), 4, 1, (5, 1), A, 3), (2, 3), 4,3), , 3), (A, 5), (2,5), (3, 5) and rad a s. The probability that the second digit drawn is odd = 35=3- (iii) There are six cases in which both the digits drawn are odd, viz., (1, 3), (1,5),8,),,5),6,1)and 6,3). ++ The probability that both the digits drawn are odd = $= #4. (8) @) Numbers (out of the first 25 numerals) which are multiples of 5 are 5, 10, 15, 20 and 25, ie, 5in all and the numbers which are multiples of 7 are 7, 14 and 21, t¢.,3 in all. Hence required number of favourable cases are 5 +3=8 + Required probability = 5 (ii) Numbers (among the first 25 numerals) which are multiples of 3 are 3, 6,9, 12, 15,18, 21, 24, ie, fin all; and the numbers, which are multiples of 7 are 7, 14,21, ie,,3 'n all. Since the number 21 is common in both the cases, the required number of distinct favourable cases is 8 + 3—1 = 10. Required probability =42 =2. 5 3-4. (¥ Four canis are drawn at random from « pack of 52 cards, Find the privity that (0) They area king, a queen, a jack and an ace. (ii) Two are kings and too are queens. iti) Two are black and lwo are red. (i) There are two cards of hearts and two cards of diamonds. 2) In shuffling a pack of cards, four are accidentally dropped, find the chance that the _missing cards should be ome from each suit. Solution. Four cards can be drawa from a well-shuffled pack of 52 cards in "Cy ‘ways, which gives the exhaustive number of cases. (i) Lking can be drawn out of the 4 kings in 4C, ways. Similarly, 1 queen, 1 jack and an ace can each be drawn in 4C; = 4 ways. Since any one of the ways of drawing a king can be associated with any one of the ways of drawing a queen, a jack and an ace, the favourable number of cases are *C; x4C x4, x4C, . Hence the required probaly =‘SuS*CyAC **6, 256 “a Required probability =“O.<°S. (iit) Since there are 26 black cards (of spades and clubs) and 26 red cards (of diamonds and hearts) in a pack of cards, the required probability = "G%"C BQ x8Cy ae @) Required probability =sHEORY OF PROBABILITY sag In all the above cases, ‘the probabilities (chances) ‘of occurrence of the orresponding events are assigned Py bat viduals and are based on their persons! agement, wisdom, intuition and Yoxpertise. These probabilities are called the peti probbiitis and represent the CoS ee ‘of betiet and the confidence, one has in aoncerrcrence of the respective event. Such * obabiites can be calculated in al he re though the probabilities assigned by Fiferent persons to the same event may cary from person t0 person: 5.7, MATHEMATICAL TOOLS : PRELIMINARY NOTIONS OF SETS “The set theory was developed by the German ‘mathematician, G. Cantor (1845 1918). 137.1. Sets and Elements of Sets. A set is 2 well-defined collection or aggregate ofall possible objects having given Proper ‘and specified according to a well” vlafinedt rule. The objects comprising a $4 6F * led clements, members or points of sey Sets are often denoted by capital letters %° WALB, C, ete. IFx is an element of the et. we write symbolically x A (x Belongs ‘AY. Ix is not a member of the set et eritex ¢ A (e does not belong to A) Sete me ‘often described by describing the Sroperties possessed by their members A etine set A of all non-negative rational aor with square less than 2 will be written 35° ‘A = (x: rational, x20, <2). rrevery clement of the set A belongs 19 the set B, ig, ifxe A= xe B then we Sty shat ed subset of B and. wate symbolically ‘Ae B(A is contained in B) oF PA (B contains A), Two sets A and B are sid to be equal or identical if A&B and BCA and write A= Bor B=A. ‘A mull or an emply set is one which does not conta any element at all and is denoted by 6 ‘Remarks. 1. Every set isa subset of itself 2. Anempty set is subset of every se + per contnning ony one element is conceptually qo from the element itself, but sale pepresented by te same sy for the sak of convenience Be le the casein all our applications of et EES ‘especially to: probability theory, wo shan have a fixed set S, (ay). given in advance, meet we shall be concerned only with we sal ve given et. The underiying set S may vary fram O08 ‘pplication to another tl seg a tras universal tof each particular disso er'9. Operations on Sets. The tion of twe Bit sets A-and B, denoted by Oa Gefined as a set consisting of all those pointy which belong to either 4 or I or both. Thus symbolically : AUB = (xixe Aorxe Bl. similarly & ‘Apel xix Aj forat least one i= 1,2, oh ‘The intersection of two sets A and B, denoted by Ao B, is defined as a set consisting of those: ‘Zlements which belong to both A and ‘B. Thus: AOB =\xixe Aand xe B} similarly 0, A = see Ajfor all i=1, 2,000 1h For example, if A = {1,2,5,8-10) and B= [2,4,8, 121, then AUB $(12,4,5,8,10, 12 and A B= (28h3:20 FUNDAMENTALS OF MATHEMATICAL STATI THEORY OF If A and B have no common point, ic., A> B=4, then the sets A and B are said ‘Thus lim. be disjoint, mutually exclusive or non-cverlapping. The relative difference of a set A from another set B, denoted by A ~B is defined 4 set consisting of those elements of A which do not belong to B. Symbolically A-B=(x:xe Aand xe B) The complement or negative of any set A, denoted by A is a set containing elements of the universal set S (say) that are not elements of A, ic., A = -A 37-3, Algebra of Sets. Now we state certain important properties con ‘operations on sets. If A, B and C are the subsets of a universal set S, then the follo\ Jaws hold : Commutative Laws : AUB =BUA, AOB=BOA ‘ Associative Law 5 @UB)UC =AUBLO nor ae (ANB) =ANBAC) Fol Distributive Law § ANBYO =(AnBUANG Sie ALBA) =ALB)AULO Complementary Law : AVA =S, ANA =6 AUS=S, and ANS=A (oA 8), AUG=A, Ano=6. Diference Law : dob eAnk A-B =A~(AMB)=(AUB)-B we shall A-(B-Q) =(A-B)U(A-C) A ting Ri (AuB)-C =(4-Qu(B-O ‘unions’ and! A-(BUQ =(A-B)n(A-C) ie, ifAe Ri (ANB)U(A-B)=A, (AM B)A\A~B)=6 Obviously ¢ A field F (r —— 22 formation of and (ANB) =A UB () Ae - Ac-ting C (A) =A Idempotency lm =: = AUA=A, ANA®A unions, 37-4. Limit of Sequence of Sets. Let (Ay) be a sequence of sets in S. The Ascfield supremum or limit superior of the sequence, usually written as lim sup A, is the fouriations all those elements which belong to A,, for infinitely many mn. Thus im sup A, = (xx € A, for infinitely many 1) ‘The set ofall those elements which belong to 4, for all but a finite number of called limit infimimuam or limit inferior of the sequence and is denoted by lim inf Ay.THEORY OF PROBABILITY! 3-21 ‘Thus lim inf A, = (xe A, forall but a finite number of 1} 3a) “The sequence {Ay} is said to have a limit ifand only if lim sup Ay = lim inf Ay and. this common value gives the limit of the sequence. Theorem 3-4. lin sup Ay = a, ( & An) and tim inf y= ©, A A) Def. {A,} is a monotone (infinite) sequence of sets if either (i) Ay C Ansa Yom OF (Ay DAns1 Ve In the former case the sequence {Ap] is said to be non-decreasing (or increasing) sequence and is usually expressed as Ay t and in the latter case it is said to be non- increasing (or decreasing) sequence and is expressed as An 4 For a monotone sequence (non-increasing or non-decreasing), the limit always exists and we have: S, Ay im case (i), fe, An T lim A, ={ "2" ne a A, in case (ii), ie, Ay /A7S. Classes of Sets. A group of sets will be termed as a class (of sets). Below we shall define some useful types of classes. ‘A ting Ris a non-empty class of sets which is closed under the formation of finite ‘unions’ and ‘difference’, ie, ifAe R,Be RthenAvBe Rand A-Be R. ‘Obviously ¢ is a member of every ring. [A field F (or an algebra) is a non-empty class of sets which is closed under the formation of finite unions and under complementation. Thus @ Ack BeF + AUBeF and (ACF > Aer. ‘Acc-ting C is a non-empty class of sets which is closed under the formation of ‘countable unions’ and ‘difference’. Thus DAaGieldn @ DAeC W AcGBEC = ABE More precisely o-ring isa ting which is closed under the formation of countable unions, “A gvfield (or o-algebra) B is a non-empty class of sets that is closed under the formation of ‘countable unions’ and complementations, ie, @ AeB, rfieta may also be defined as a field which is closed under the formation of countable unions. y2n, = UA eR BACB = Aes. A3.28 FUNDAMENTALS OF MATHEMATICAL STATISI x Since, P(S) = & P(e) =1, on using G4), we get it P(e) epi f= V2 N in the case of natural assignment of probabilities. For example, inthe case of random toss of a coin, = (HT) and the nat assignment of probabilities gives: P (H) = P(T) L Similarly, ina randiom throw of a i, $= {2 2,» 6)= {e/a fe (889) natural assignment of probabilities is P(e) = 71= 1/200 6 In the problems relating to randam throw%s) of coin(s) oF die (dice), the assignment of probabilities is possible only if the coins/ dice are (i unifor aa eed, Similarly, in drawing cards from a wellshuifled pack of cards, for Sesignment of probabilities, all the draws must be made at random. vr the following discussion, whenever we use the adjective such as : lt ‘fair or umbinet dic’, card or arise dravin at "random el we wht sure Fer oe inssignment is understood, Also if no assignment of probabilities i Bh ‘ny question, the natural assignment is to be taken. 338.5, Axiomatic Probability. A purely mathematical definition of probs cannot give us the actual value of P(A), the probability of occurrence of thee sani hig mnast be considered as a function defined on all events. Accordingly, a8 ara of any function we need a domain space which is the o-field B of Ihe fgenerated by S za range space which isthe closed interval [0 1] on the real line, Bele which gesigns a value to every element of the domain space B, Definition. Probability Function. P(A) is the probability function defined om Feld B of events if the following properties ar axioms ld 41. Foreach A « B, P(A) is defined, is real and P(A) 20. (Axiom of non-negativity) 2P(S)=1 (Axiom of certainty) 23. (Al is any finite or infinite sequence of disjoint exerts in B, then ) “An obvious advantage of this definition is that it is not concerned with ass numerical values to the probabilities of events. ‘Remarks 1. The set function P defined on o-fild B, taking its vals in the rea satisfying the above three axioms is called the probability meusure f The same definition of probability applies to countable sample space except that restrictions must be placed on S and its subsets, It important to realise that for 969 TTesenption ofa probability mensure three things must be apeified i the sample 5 = 2% Pap (Axiom of additivity) P( OA ae i gba) B formed from certain subsets of Sand st function P. The triplet (6 erie sled the probity space. In most elementary applications $s finite and the o-al fs taken to be the collection of all subsets ofS. + Clussical Probability as a Special Case of Axiomatic Probability. Let us ov random experiment whose sample space Sas finite numberof points, say N, given BYTHEORY OF PROBABILITY 3-29 SH le ty stale where obviously ¢, c,....¢y are mutually disjoint events. Let us now suppose that all the outcomes of the random experiment are equally likely so ‘hat the probability of each of the singleton events (¢ is 1/N, 12, Pled= yi AEN 2 ooo ag. Further, let an event A consist of k distinct point of S, say ¢,¢q ».¢q. Then by the axiom of additivity (Axiom 3), we get > PUA) =X Plel= Beh [Using (9) ‘Numberofsample pointsin _ n(A) Lez * Number of sample points in S = nS) = Number of cases favourable tothe event A, ‘xhausbve number of cases hich isthe result obtained by the classical definition of probability. Hence, the classical probability may be regarded as a special ease of axiomatic probability 4. It is interesting to see that there are some formal statements of the properties of events derived from the frequency approach. ‘From axiomatic definition of probability, we get, P (A) = md} 20, as in Axiom 1 Alo since nS) =, we haveon using (37, (5) =! a5 in Axion? Further inthe cae of two mutunlyexlisive (or disjoint) events A and B defined by sample points n(A) and (8), the number of sample points belonging to AUB is n(A) + n(f). PAD) = BALE ne 2H = 2 + 8D. D(A) + PCB), as in Axiom 3. Hence the statistical or frequency approach of probability may be regarded as a special ‘ase of axiomatic probability. 38-6, Algebra of Events, For events A, B,C (@ AUB={|we S:we Aorwe B} Gi) ANB={we S:we Aandwe B} (iii) A (A complement) = (we $ :w¢ A} (i) A-B= (me $:e Abutae B} (©) Similar generalisations for 0, Ay, A, Avete. () ACB = — foreverywe Awe B, (vi) B5A4A cB. (vil) A= Bifand only if A and B have same elements, ie,,ifA Band BCA. (ix) A and B disjoint (mutually exclusive) => A cy 8 = 6 (null set) () AUB canbe denoted by A+ BifA and B are disjoint: (vi) AAB denotes those w belonging to exactly one of A and B, i, ASB = ABUAB= AB + AB (disjoint events) ‘Remark. Since the events are subsets ofS, al the laws of set theory, viz, commutative laws, associative laws, distributive laws, De-Morgan's law, etc, hold for algebra of events,3-30 FUNDAMENTALS OF MATHEMATICAL STATISTICS! ‘TABLE-GLOSSARY OF PROBABILITY TERMS Statement ing Lem of seo 1. Atleast one ofthe events A or B occurs oe AUB 2 Both the events A and B occur eeAae 3. Neither A nor B occurs oe And 4. Event A occurs and B does not occu. weAn E 5. Exactly one of the events Aor B occurs oe AaB ‘6 Not more than one of the events A or B we (40 B)U(An BUA no B) 7. event A occurs, s0 does B. AcB 8. Events A and B are mutually exclusive, AnB=o 9, Complementary event of A. ic 10. Sample space Universal set Example 3-23. A, B and C are three arbitrary events. Find expression for the events noted below, in the context of A, Band C: (i) Only A occurs, (ii) Both A and B, but not C, occur, (ii) Al tree events occur, (iv) Atleast one occurs, (0) At least two aceur, (oi) One and no more occurs, (oii) Two and no more occur, (citi) None occurs. Solution. @ AnB aE, WANBAC, (iiANBAC (iv)AUBUC, @ ANB C) LANE AQU(ARBAQUIANBAg @) AnBa OC) ulA aba T)uan FAQ (i) ANBAT) UANBAQULA A Fag (itt) ANB a C) or AGBUC 3.9, SOME THEOREMS ON PROBABILITY In this section, we shall prove a few simple theorems which help us evaluate the probabilities of some complicated events in a rather simple way. In proving these theorems, we shall follow the axiomatic approach, based on the three axioms, discussed in § 385. Theorem 3-2. Probability ofthe impossible event is zero, ic., P(@) =0. 38) Proof, Impossible event contains no sample point and hence the certain event S and the impossible event are mutually exclusive, 2 Sud=5 = PSU d)=P(S) Hence, using Axiom 2 of probability, ie., Axiom of Additivity, we get P(S)+P@)=P(S) = P()=0 ‘Remark. It may be noted P(A) = 0, does not imply that A is necessarily an empty set. In practice, probability ‘is assigned to the events which are so rare that they happen only once in a lifetime. For example, if a person who does not know typing is asked to type one page of the ‘manuscript of a book, the probability of the event that be will type it correctly without any mistake is 0,a THEORY OF PROBABILITY—! 331 “Av another iusraion, let ws consider the random tssing of cin, The event that the coin — will stand erect on its edge i assigned the probability 0. “The study of continuous random variable provides another illustration ts (6 that P(A) <0, Meas not imply A =6, beens in ase of continuous random variable, the probability ata point in always 2x0, ic, PX =) =0 (See Chapter Sh ‘Theorem 3-3 Probability ofthe complementary ecent of A is given by P(A )=1-P(A). 9) Proof, A andA are mutually disjoint events, so that AUAsS 3 P(AVA)=P(S) Hence, from Axioms 2 and 3 of probability, we have PiayeP(A)=P(S)=1 => PCA) =1=P(A) Cor. 1.Wehave P(A)=1-P(A)S1 [- P(A) 20, by Axiom 1] = Further, since P(A) 20 (Axiom 1) a 5 0s P(A) S1. (890) Cor. 2. P(4)=0,since ¢ =5 and P(*)=P(S )=1-P(S) =1-1=0. Theorem 3-4, For any two events A and B, se have (P(A 9 B)=P@)-P (ANB) (iP (ANB) =P(A)-P(ANB) Proof, From the Venn diagram, we get Ba(AnB)U(A OB) where A Band AmB are disjoint events. Hence by Axiom (3), we get P(B) =P(ANB)+P (A 08) = P(A 9B) =PB)-P ANB) (10) (i Similarly, we have Az(ANB)UANB), mail where (A 0B) and A.B are disjoint events. Hence, by Axiom 3, we get —s P(A)=P (ANB)+P(ANB) = P (AaB) P(AMB) - Theorem 3-5.1fB 0, P(A 1B)S P(A) Proof. n(ANB)sn(A) and (B)