Class 12 PCM . Mathematics . Mathematics Part II . Probability

Chapter 13: Probability

Ch 13

Mathematics

Class 12 PCM

Lecture 1

Introduction

Reference video · Lecture 1Introduction

Chapter focus and assumptions

What this chapter develops

  • Starting point: Probability measures uncertainty in the outcomes of a random experiment.
  • This lecture: Introduces conditional probability—probability calculated when another event is known to have occurred.
  • Later ideas: Conditional probability supports Bayes’ theorem, the multiplication rule, and independence; the chapter also covers random variables, probability distributions, mean, variance, and the binomial distribution.
  • Default assumption: Outcomes are treated as equally likely unless stated otherwise.

Conditional probability

Definition

Conditional probability — The conditional probability of event EE given that event FF has occurred is the probability of EE calculated within the restricted sample space FF.

Conditional probability formula

For events EE and FF in the same sample space, the outcomes counted in the restricted space FF that also satisfy EE are exactly those in E∩FE\cap F. The formula requires P(F)>0P(F)>0.

P(E∣F)=P(E∩F)P(F),P(F)>0P(E\mid F)=\frac{P(E\cap F)}{P(F)},\quad P(F)>0

Key points

  • Restricted space: Once FF is known to have occurred, outcomes outside FF are excluded.
  • Favourable outcomes: The outcomes in FF that also make EE occur form E∩FE\cap F.
  • Equally likely outcomes: P(E∣F)=n(E∩F)n(F)P(E\mid F)=\dfrac{n(E\cap F)}{n(F)}, provided n(F)>0n(F)>0.
  • Condition: If P(F)=0P(F)=0, the conditional probability P(E∣F)P(E\mid F) is not defined by this formula.
Conditional probability
P(E∣F)=P(E∩F)P(F),P(F)>0P(E\mid F)=\frac{P(E\cap F)}{P(F)},\quad P(F)>0

Use when finding the probability of EE after learning that FF has occurred.

Finding a conditional probability

  1. Identify the given event FF; restrict attention to outcomes in FF.
    FF
  2. Find the outcomes common to EE and FF, namely E∩FE\cap F.
    E∩FE\cap F
  3. Divide the probability of the common outcomes by the probability of the given event; check that P(F)>0P(F)>0.
    P(E∣F)=P(E∩F)P(F)P(E\mid F)=\frac{P(E\cap F)}{P(F)}
Worked example

Find the probability of at least two heads given that the first coin shows a tail.

  1. Three fair coins are tossed. Let EE be the event of at least two heads, and FF the event that the first coin shows a tail.
    S={HHH,HHT,HTH,THH,HTT,THT,TTH,TTT}S=\{HHH,HHT,HTH,THH,HTT,THT,TTH,TTT\}
  2. List the outcomes in each event.
    E={HHH,HHT,HTH,THH},F={THH,THT,TTH,TTT}E=\{HHH,HHT,HTH,THH\},\quad F=\{THH,THT,TTH,TTT\}
  3. The only outcome common to both events is THHTHH; hence find the two required probabilities.
    E∩F={THH},P(E∩F)=18,P(F)=48E\cap F=\{THH\},\quad P(E\cap F)=\frac{1}{8},\quad P(F)=\frac{4}{8}
  4. Apply the conditional probability formula.
    P(E∣F)=P(E∩F)P(F)=1/84/8=14P(E\mid F)=\frac{P(E\cap F)}{P(F)}=\frac{1/8}{4/8}=\frac{1}{4}
Answer: P(E∣F)=14P(E\mid F)=\frac{1}{4}.
⚠ Common mistake

Do not divide by P(E)P(E): When the condition is that FF occurred, the denominator is P(F)P(F); the restricted sample space is FF.

Practice

A card is drawn from a standard 52-card deck. Given that it is red, what is the probability that it is a king?

Hint: There are 26 red cards, of which 2 are kings.
Key step: Let KK denote drawing a king and RR denote drawing a red card. Use P(K∣R)=n(K∩R)n(R)P(K\mid R)=\dfrac{n(K\cap R)}{n(R)}.
Solution: P(K∣R)=226=113P(K\mid R)=\frac{2}{26}=\frac{1}{13}.
Practice

If P(A∩B)=0.18P(A\cap B)=0.18 and P(B)=0.6P(B)=0.6, find P(A∣B)P(A\mid B).

Hint: Divide the intersection probability by the probability of the given event.
Key step: Substitute into P(A∣B)=P(A∩B)P(B)P(A\mid B)=\dfrac{P(A\cap B)}{P(B)}.
Solution: P(A∣B)=0.180.6=0.3P(A\mid B)=\frac{0.18}{0.6}=0.3.

Quick recap

  • Conditional probability is computed after restricting the sample space to the event already known to have occurred.
  • The outcomes favourable to both events are represented by E∩FE\cap F.
  • Use P(E∣F)=P(E∩F)P(F)P(E\mid F)=\dfrac{P(E\cap F)}{P(F)} when P(F)>0P(F)>0.
  • For equally likely outcomes, use P(E∣F)=n(E∩F)n(F)P(E\mid F)=\dfrac{n(E\cap F)}{n(F)}.
  • The condition event FF determines the denominator.
Lecture 2

Properties Of Conditional Probability

Reference video · Lecture 2Properties Of Conditional Probability

Conditional Probability and Its Properties

Definition

Conditional probability — The probability of an event AA given that event BB has occurred, where P(B)>0P(B)>0.

Conditional probability of the sample space

For any event FF with P(F)>0P(F)>0, the sample space and the given event are certain under the condition that FF occurred: P(S∣F)=P(F∣F)=1P(S|F)=P(F|F)=1.

Addition rule

For events A,BA,B and FF with P(F)>0P(F)>0, P((A∪B)∣F)=P(A∣F)+P(B∣F)−P((A∩B)∣F)P((A\cup B)|F)=P(A|F)+P(B|F)-P((A\cap B)|F). If AA and BB are disjoint, the intersection term is zero.

Complement rule

For any event EE and FF with P(F)>0P(F)>0, the conditional probabilities of EE and its complement add to 11: P(E′∣F)=1−P(E∣F)P(E'|F)=1-P(E|F).

Conditional probability
P(A∣B)=P(A∩B)P(B),P(B)>0P(A|B)=\frac{P(A\cap B)}{P(B)},\quad P(B)>0

Use this definition for both equally likely and unequally likely outcomes.

Conditional probability for equally likely outcomes
P(A∣B)=n(A∩B)n(B)P(A|B)=\frac{n(A\cap B)}{n(B)}

When outcomes in the restricted event BB are equally likely, count outcomes within BB.

Finding a conditional probability

  1. Identify the event whose probability is required, AA, and the given event, BB.
    A, BA,\ B
  2. Find the outcomes in both events, A∩BA\cap B.
    A∩BA\cap B
  3. Apply the conditional probability formula; for equally likely outcomes, count within BB.
    P(A∣B)=P(A∩B)P(B)P(A|B)=\frac{P(A\cap B)}{P(B)}
Worked example

A family has two children. What is the probability that both are boys given that at least one is a boy?

  1. A family has two children. Find the probability that both are boys, given that at least one is a boy. Let EE be both boys and FF be at least one boy.
    E={(b,b)},F={(b,b),(g,b),(b,g)}E=\{(b,b)\},\quad F=\{(b,b),(g,b),(b,g)\}
  2. The outcomes in the given event FF are equally likely, and only one is in EE.
    E∩F={(b,b)},n(F)=3,n(E∩F)=1E\cap F=\{(b,b)\},\quad n(F)=3,\quad n(E\cap F)=1
  3. Apply the conditional counting formula.
    P(E∣F)=n(E∩F)n(F)=13P(E|F)=\frac{n(E\cap F)}{n(F)}=\frac{1}{3}
Answer: 13\frac{1}{3}
⚠ Common mistake

Do not use the full sample space after a condition is given. Restrict the outcomes to the given event first; also check that its probability is non-zero.

Conditional Probability with Unequally Likely Outcomes

Weighted outcomes

  • General case: The conditional probability formula still applies when elementary outcomes are not equally likely; use their assigned probabilities.
  • Two-stage experiment: If a coin shows heads, toss it again; if it shows tails, roll a die. Outcomes are HH,HT,T1,…,T6HH, HT, T1,\ldots,T6.
  • Outcome probabilities: P(HH)=P(HT)=14P(HH)=P(HT)=\frac14; each outcome T1T1 through T6T6 has probability 112\frac1{12}.
  • Given at least one tail: The condition has probability P(F)=14+6(112)=34P(F)=\frac14+6\left(\frac1{12}\right)=\frac34.
  • Die result greater than 44 and at least one tail: P(E∩F)=P(T5)+P(T6)=16P(E\cap F)=P(T5)+P(T6)=\frac16; hence P(E∣F)=29P(E|F)=\frac29.
Tree diagram of the outcomes when a coin result determines whether to toss again or roll a die.
Tree diagram of the outcomes when a coin result determines whether to toss again or roll a die.
Probability tree assigning probabilities to the possible outcomes of the two-stage experiment.
Probability tree assigning probabilities to the possible outcomes of the two-stage experiment.
Practice

A card numbered from 11 to 1010 is drawn. Given that its number is greater than 33, find the probability that it is even.

Hint: Restrict the possible numbers to those greater than 33.
Key step: There are 77 eligible numbers, of which 44 are even.
Solution: The eligible set is {4,5,6,7,8,9,10}\{4,5,6,7,8,9,10\}; its even members are {4,6,8,10}\{4,6,8,10\}. Thus P(even∣>3)=47P(\text{even}|{>}3)=\frac47.
Practice

A die is thrown twice, and the sum is known to be 66. Find the conditional probability that a 44 appeared at least once.

Hint: List only ordered pairs whose sum is 66.
Key step: The possible pairs are (1,5),(2,4),(3,3),(4,2),(5,1)(1,5),(2,4),(3,3),(4,2),(5,1); two contain a 44.
Solution: Among the 55 equally likely pairs given the sum, 22 contain a 44. Therefore, the probability is 25\frac25.
Practice

A coin is tossed. If it shows heads, toss it again; if it shows tails, roll a die. Find the probability that the die shows more than 44, given that at least one tail occurred.

Hint: Use the probabilities of the outcomes, not a simple count of outcomes.
Key step: P(F)=34P(F)=\frac34 and P(E∩F)=16P(E\cap F)=\frac16.
Solution: The two qualifying outcomes are T5T5 and T6T6, each with probability 112\frac1{12}. Thus P(E∣F)=1/63/4=29P(E|F)=\frac{1/6}{3/4}=\frac29.

Quick recap

  • Conditional probability is P(A∣B)=P(A∩B)P(B)P(A|B)=\frac{P(A\cap B)}{P(B)}, provided P(B)>0P(B)>0.
  • Given FF, both the sample space and FF have conditional probability 11.
  • The conditional addition rule subtracts the overlap; for disjoint events, add the probabilities directly.
  • The conditional complement rule is P(E′∣F)=1−P(E∣F)P(E'|F)=1-P(E|F).
  • For equally likely outcomes, count favourable outcomes within the condition; otherwise use their actual probabilities.
  • Always restrict the sample space to the given condition before calculating.
Lecture 3

Multiplication Theorem on Probability

Reference video · Lecture 3Multiplication Theorem on Probability

Multiplication Theorem

Multiplication theorem for two events

For events EE and FF, P(E∩F)=P(E)P(F∣E)=P(F)P(E∣F)P(E \cap F)=P(E)P(F \mid E)=P(F)P(E \mid F), using the conditional-probability form when the conditioning event has nonzero probability.

P(E∩F)=P(E)P(F∣E)=P(F)P(E∣F)P(E \cap F)=P(E)P(F \mid E)=P(F)P(E \mid F)
Three or more events
P(E∩F∩G)=P(E)P(F∣E)P(G∣E∩F)P(E \cap F \cap G)=P(E)P(F \mid E)P(G \mid E \cap F)

For three events, multiply each probability by the next event's probability conditional on all previous events.

Extension to n events
P(⋂i=1nEi)=P(E1)∏i=2nP(Ei∣⋂j=1i−1Ej)P\left(\bigcap_{i=1}^{n} E_i\right)=P(E_1)\prod_{i=2}^{n}P\left(E_i \mid \bigcap_{j=1}^{i-1}E_j\right)

Continue conditioning each event on the occurrence of all events before it.

Using the multiplication theorem

  1. Name the events in the order they occur.
  2. Find the probability of the first event.
    P(E1)P(E_1)
  3. Find each later probability conditional on the earlier events.
    P(Ei∣E1∩⋯∩Ei−1)P(E_i \mid E_1 \cap \cdots \cap E_{i-1})
  4. Multiply the probabilities to obtain the probability of their joint occurrence.
Worked example

What is the probability that both drawn balls are black?

  1. An urn has 10 black and 5 white balls. Two balls are drawn without replacement. Find the probability that both are black.
  2. The probability that the first ball is black is:
    P(E)=1015P(E)=\frac{10}{15}
  3. After a black ball is drawn, 9 black balls remain among 14 balls:
    P(F∣E)=914P(F \mid E)=\frac{9}{14}
  4. Apply the multiplication theorem:
    P(E∩F)=1015×914=37P(E \cap F)=\frac{10}{15}\times\frac{9}{14}=\frac{3}{7}
Answer: 37\frac{3}{7}
⚠ Common mistake

Do not use the first-draw probability again for a draw without replacement; the conditional probability changes because the contents of the urn change.

Practice

Three cards are drawn successively without replacement from a 52-card deck. Find the probability that the first two are kings and the third is an ace.

Hint: Use the conditional probability for each draw, updating the deck after each king.
Key step: 452×351×450\frac{4}{52}\times\frac{3}{51}\times\frac{4}{50}
Solution: P=452×351×450=25525P=\frac{4}{52}\times\frac{3}{51}\times\frac{4}{50}=\frac{2}{5525}.

Independent Events

Independence test
P(E∩F)=P(E)P(F)P(E \cap F)=P(E)P(F)

Use this test to decide whether two events are independent.

Independent vs mutually exclusive
PointIndependentMutually exclusive
MeaningOne event does not affect the probability of the other.The events have no outcomes in common.
IntersectionMay be nonempty.E∩F=∅E \cap F=\varnothing and P(E∩F)=0P(E \cap F)=0.
With positive probabilitiesCannot be mutually exclusive.Cannot be independent.
Mutual independence of three events

Events A,B,CA,B,C are mutually independent only if all three pairs are independent and the joint probability is also the product of all three individual probabilities.

At least one of two independent events
P(A∪B)=1−P(A′)P(B′)=P(A)+P(B)−P(A)P(B)P(A \cup B)=1-P(A')P(B')=P(A)+P(B)-P(A)P(B)

For independent AA and BB, use the complement event that neither occurs.

Complements of independent events

If EE and FF are independent, then EE and F′F', E′E' and FF, and E′E' and F′F' are also independent.

P(E∩F′)=P(E)P(F′)P(E \cap F')=P(E)P(F')

Testing independence

  1. Find the probabilities of each event.
    P(E),P(F)P(E),\quad P(F)
  2. Find the probability that both events occur.
    P(E∩F)P(E \cap F)
  3. Compare the joint probability with the product of the individual probabilities.
    P(E∩F)andP(E)P(F)P(E \cap F)\quad\text{and}\quad P(E)P(F)
  4. Equal values mean independent events; unequal values mean dependent events.
Worked example

Which of the pairs (E,F)(E,F), (E,G)(E,G) and (F,G)(F,G) are independent?

  1. Three coins are tossed. Let EE be three heads or three tails, FF be at least two heads, and GG be at most two heads. Determine which pairs are independent.
  2. There are 8 equally likely outcomes. The event probabilities are:
    P(E)=28=14,P(F)=48=12,P(G)=78P(E)=\frac{2}{8}=\frac14,\quad P(F)=\frac{4}{8}=\frac12,\quad P(G)=\frac{7}{8}
  3. For EE and FF, their intersection contains only HHH, so the joint probability equals the product:
    P(E∩F)=18=P(E)P(F)=14×12P(E \cap F)=\frac18=P(E)P(F)=\frac14\times\frac12
  4. For EE and GG, their intersection contains only TTT, so the probabilities are unequal:
    P(E∩G)=18≠P(E)P(G)=732P(E \cap G)=\frac18\ne P(E)P(G)=\frac{7}{32}
  5. For FF and GG, the intersection has 3 outcomes, so the probabilities are unequal:
    P(F∩G)=38≠P(F)P(G)=716P(F \cap G)=\frac38\ne P(F)P(G)=\frac{7}{16}
Answer: (E,F)(E,F) is independent; (E,G)(E,G) and (F,G)(F,G) are dependent.
Practice

A die is thrown once. Let EE be the event of getting a multiple of 3 and FF the event of getting an even number. Are the events independent?

Hint: Compare P(E∩F)P(E \cap F) with P(E)P(F)P(E)P(F).
Key step: E={3,6}, F={2,4,6}, E∩F={6}E=\{3,6\},\ F=\{2,4,6\},\ E\cap F=\{6\}
Solution: P(E)=13P(E)=\frac13, P(F)=12P(F)=\frac12, and P(E∩F)=16=P(E)P(F)P(E\cap F)=\frac16=P(E)P(F); hence the events are independent.
Practice

An unbiased die is thrown twice. Let AA mean an odd result on the first throw and BB an odd result on the second. Check independence.

Hint: There are 36 equally likely ordered outcomes.
Key step: P(A)=P(B)=12,P(A∩B)=936P(A)=P(B)=\frac12,\quad P(A\cap B)=\frac{9}{36}
Solution: P(A∩B)=14=P(A)P(B)P(A\cap B)=\frac14=P(A)P(B); hence AA and BB are independent.
Practice

If EE and FF are independent, show that EE and F′F' are independent.

Hint: Use E=(E∩F)∪(E∩F′)E=(E\cap F)\cup(E\cap F'); the two parts are disjoint.
Key step: P(E∩F′)=P(E)−P(E∩F)P(E\cap F')=P(E)-P(E\cap F)
Solution: P(E∩F′)=P(E)−P(E)P(F)=P(E)(1−P(F))=P(E)P(F′)P(E\cap F')=P(E)-P(E)P(F)=P(E)(1-P(F))=P(E)P(F'). Therefore EE and F′F' are independent.
⚠ Common mistake

Do not confuse independence with mutual exclusivity: mutual exclusivity concerns shared outcomes, while independence concerns probabilities. If both events have nonzero probabilities, mutually exclusive events are dependent.

Quick recap

  • For two events, P(E∩F)=P(E)P(F∣E)=P(F)P(E∣F)P(E\cap F)=P(E)P(F\mid E)=P(F)P(E\mid F).
  • For successive events, multiply each probability conditional on all previous events.
  • Two events are independent exactly when P(E∩F)=P(E)P(F)P(E\cap F)=P(E)P(F).
  • For three mutually independent events, test all three pairs and the triple intersection.
  • Independence is different from mutual exclusivity; positive-probability mutually exclusive events are dependent.
  • If two events are independent, their complements are also independent with the corresponding events.
  • For independent AA and BB, P(A∪B)=1−P(A′)P(B′)P(A\cup B)=1-P(A')P(B').
Lecture 4

Bayes' Theorem

Reference video · Lecture 4Bayes' Theorem

Partition of a Sample Space

  • Partition: Events form a partition of the sample space when they are pairwise disjoint, exhaustive, and each has nonzero probability.
  • Example: Any nonempty event and its complement form a partition of the sample space.

Theorem of Total Probability

Total probability
P(A)=∑j=1nP(Ej)P(A∣Ej)P(A)=\sum_{j=1}^{n}P(E_j)P(A\mid E_j)

Use to find the overall probability of an event across all partition cases.

Event A overlaps the partition regions E_1, E_2 and E_3.
Event AA overlaps the partition regions E1E_1, E2E_2 and E3E_3.

Bayes’ Theorem

Bayes’ theorem

For a partition E1,…,EnE_1,\ldots,E_n and an event AA with P(A)>0P(A)>0, the conditional probability of hypothesis EiE_i given AA equals its prior-weighted likelihood divided by the total probability of AA.

P(Ei∣A)=P(Ei)P(A∣Ei)∑j=1nP(Ej)P(A∣Ej)P(E_i\mid A)=\frac{P(E_i)P(A\mid E_i)}{\sum_{j=1}^{n}P(E_j)P(A\mid E_j)}

Applying Bayes’ theorem

  1. List the possible causes as a partition and identify the observed event.
  2. Write each cause’s prior probability and the probability of the observation under that cause.
  3. Multiply each prior by its corresponding likelihood and add the products to get the total probability of the observation.
    P(Ej)P(A∣Ej)P(E_j)P(A\mid E_j)
  4. Divide the product for the required cause by the total probability of the observation.
    P(Ei∣A)=P(Ei)P(A∣Ei)P(A)P(E_i\mid A)=\frac{P(E_i)P(A\mid E_i)}{P(A)}
Worked example

What is the probability that a defective bolt was manufactured by machine B?

  1. Machines A, B and C make 25%25\%, 35%35\% and 40%40\% of a factory’s bolts. Their defective rates are 5%5\%, 4%4\% and 2%2\%. Given that a bolt is defective, find the probability it came from B.
  2. Let B1,B2,B3B_1,B_2,B_3 denote production by A, B and C, and let DD denote a defective bolt.
    P(B1)=0.25,P(B2)=0.35,P(B3)=0.40P(B_1)=0.25,\quad P(B_2)=0.35,\quad P(B_3)=0.40
  3. Apply Bayes’ theorem to the hypothesis that machine B made the bolt.
    P(B2∣D)=0.35(0.04)0.25(0.05)+0.35(0.04)+0.40(0.02)P(B_2\mid D)=\frac{0.35(0.04)}{0.25(0.05)+0.35(0.04)+0.40(0.02)}
  4. Evaluate the numerator and denominator.
    P(B2∣D)=0.0140.0345=2869P(B_2\mid D)=\frac{0.014}{0.0345}=\frac{28}{69}
Answer: 2869\frac{28}{69}

Other Bayes’ theorem applications

  • Two bags: Bag I has 33 red and 44 black balls; Bag II has 55 red and 66 black balls. With equal bag priors, the probability a red ball came from Bag II is (1/2)(5/11)(1/2)(3/7)+(1/2)(5/11)=3568\frac{(1/2)(5/11)}{(1/2)(3/7)+(1/2)(5/11)}=\frac{35}{68}.
  • Three boxes: Choose one of three boxes equally: all gold, all silver, or one gold and one silver coin. Given a gold coin was drawn, the probability the other coin is gold is (1/3)(1)(1/3)(1)+(1/3)(0)+(1/3)(1/2)=23\frac{(1/3)(1)}{(1/3)(1)+(1/3)(0)+(1/3)(1/2)}=\frac{2}{3}.
  • Late doctor—transport priors: Train, bus, scooter and other transport have probabilities 310,15,110,25\frac{3}{10},\frac{1}{5},\frac{1}{10},\frac{2}{5}, respectively.
  • Late doctor—late probabilities: The probabilities of being late by train, bus, scooter and other transport are 14,13,112,0\frac{1}{4},\frac{1}{3},\frac{1}{12},0, respectively.
  • Late doctor—posterior: Given that he is late, the probability he came by train is (3/10)(1/4)(3/10)(1/4)+(1/5)(1/3)+(1/10)(1/12)+(2/5)(0)=12\frac{(3/10)(1/4)}{(3/10)(1/4)+(1/5)(1/3)+(1/10)(1/12)+(2/5)(0)}=\frac{1}{2}.
⚠ Common mistake

Do not confuse P(A∣Ei)P(A\mid E_i) (likelihood of the observation under a cause) with P(Ei∣A)P(E_i\mid A) (probability of the cause after observing it); Bayes’ theorem converts the first into the second.

Random Variables

Values of two random variables
OutcomeXX: number of headsYY: heads minus tails
HHHH2222
HTHT1100
THTH1100
TTTT00−2-2
Practice

A job has probability 0.650.65 of a strike. It finishes on time with probability 0.320.32 if there is a strike and 0.800.80 if there is no strike. Find the probability it finishes on time.

Hint: Use strike and no strike as a two-event partition.
Key step: P(A)=P(B)P(A∣B)+P(B′)P(A∣B′)P(A)=P(B)P(A\mid B)+P(B')P(A\mid B').
Solution: 0.65(0.32)+0.35(0.80)=0.4880.65(0.32)+0.35(0.80)=0.488.
Practice

A condition affects 0.1%0.1\% of a population. A test is positive for 90%90\% of affected people and for 1%1\% of unaffected people. Given a positive result, find the probability the person is affected.

Hint: Use prior probabilities 0.0010.001 and 0.9990.999, then apply Bayes’ theorem.
Key step: P(E∣A)=0.001(0.90)0.001(0.90)+0.999(0.01)P(E\mid A)=\frac{0.001(0.90)}{0.001(0.90)+0.999(0.01)}.
Solution: 0.00090.01089=901089≈0.083\frac{0.0009}{0.01089}=\frac{90}{1089}\approx 0.083.
Practice

A fair die is rolled. A person who tells the truth with probability 34\frac{3}{4} reports that the result is six. Find the probability the roll was actually six.

Hint: The report can occur after a six, or after a non-six when the person lies.
Key step: P(E∣S1)=34P(E\mid S_1)=\frac{3}{4} and P(E∣S2)=14P(E\mid S_2)=\frac{1}{4}.
Solution: 16⋅3416⋅34+56⋅14=38\frac{\frac{1}{6}\cdot\frac{3}{4}}{\frac{1}{6}\cdot\frac{3}{4}+\frac{5}{6}\cdot\frac{1}{4}}=\frac{3}{8}.

Quick recap

  • A partition consists of pairwise disjoint, exhaustive events with positive probabilities.
  • Total probability adds the contributions of every partition case that can produce the event.
  • Bayes’ theorem updates a cause’s prior probability after an observation.
  • The denominator in Bayes’ theorem is the total probability of the observed event.
  • Do not reverse conditional probabilities without applying Bayes’ theorem.
  • A random variable assigns a real number to every sample-space outcome.
Lecture 5

Miscellaneous Examples

Reference video · Lecture 5Miscellaneous Examples

Bayes’ Theorem: Identifying the Box

Definition

Bayes’ theorem — Bayes’ theorem finds the probability of a cause or source after an observed event has occurred.

Definition

Partition of a sample space — A set of events forms a partition when the events are pairwise disjoint and their union is the entire sample space.

Total probability and Bayes’ theorem

If E1,…,EnE_1,\ldots,E_n partition the sample space and P(Ej)>0P(E_j)>0, then P(A)=∑j=1nP(Ej)P(A∣Ej)P(A)=\sum_{j=1}^{n}P(E_j)P(A\mid E_j). For P(A)>0P(A)>0, P(Ei∣A)=P(Ei)P(A∣Ei)∑j=1nP(Ej)P(A∣Ej)P(E_i\mid A)=\frac{P(E_i)P(A\mid E_i)}{\sum_{j=1}^{n}P(E_j)P(A\mid E_j)}.

Balls in the four boxes
BoxBlackWhiteRedBlue
I3456
II2222
III1231
IV4315

Set up the events

  • Source events: Let EiE_i mean box ii is selected; choosing uniformly gives P(Ei)=14P(E_i)=\frac14 for each box.
  • Observed event: Let AA mean the drawn ball is black. Then P(A∣E1)=16P(A\mid E_1)=\frac16, P(A∣E2)=14P(A\mid E_2)=\frac14, P(A∣E3)=17P(A\mid E_3)=\frac17, and P(A∣E4)=413P(A\mid E_4)=\frac4{13}.
Posterior probability of box III
P(E3∣A)=P(E3)P(A∣E3)∑i=14P(Ei)P(A∣Ei)P(E_3\mid A)=\frac{P(E_3)P(A\mid E_3)}{\sum_{i=1}^{4}P(E_i)P(A\mid E_i)}

Use the probability of drawing black from each box as the likelihood, weighted by the chance of choosing that box.

Applying Bayes’ theorem

  1. Define the possible sources and the observed event.
  2. Find each source probability and the conditional probability of the observation from that source.
  3. Multiply each source probability by its likelihood; divide the target source’s product by the sum of all products.
Worked example

A box is chosen uniformly at random, and a ball drawn from it is black. Find the probability that the ball came from box III.

  1. The probability of a black ball from each box equals its black-ball count divided by its total.
    P(A∣E1)=318=16,P(A∣E2)=28=14,P(A∣E3)=17,P(A∣E4)=413P(A\mid E_1)=\frac{3}{18}=\frac16,\quad P(A\mid E_2)=\frac28=\frac14,\quad P(A\mid E_3)=\frac17,\quad P(A\mid E_4)=\frac4{13}
  2. Apply Bayes’ theorem and cancel the equal box-selection probabilities.
    P(E3∣A)=14⋅1714(16+14+17+413)=156947≈0.165P(E_3\mid A)=\frac{\frac14\cdot\frac17}{\frac14\left(\frac16+\frac14+\frac17+\frac4{13}\right)}=\frac{156}{947}\approx0.165
Answer: 156947≈0.165\frac{156}{947}\approx0.165.
⚠ Common mistake

Do not use only the likelihood P(A∣E3)P(A\mid E_3) as the answer; Bayes’ theorem must account for every possible box in the denominator.

Alternating Die Throws

Definition

Repeated-turn winning probability — When players alternate independent trials until success, a player’s possible winning events form a geometric series.

Infinite geometric series

For first term aa and common ratio rr with ∣r∣<1|r|<1, the infinite sum is a1−r\frac{a}{1-r}.

Winning probabilities when A starts
P(S)=16,P(F)=56,P(A wins)=161−(56)2=611,P(B wins)=511P(S)=\frac16,\quad P(F)=\frac56,\quad P(A\text{ wins})=\frac{\frac16}{1-(\frac56)^2}=\frac6{11},\quad P(B\text{ wins})=\frac5{11}

For a fair die, let success be rolling a six and failure be not rolling a six. Each round consists of A’s throw followed by B’s throw.

Summing repeated winning chances

  1. Find the success and failure probabilities for one turn.
  2. Write the first few ways the player can win, including the failures before their winning turn.
  3. Identify the first term and common ratio, then sum the infinite geometric series.
  4. Use the complement for the other player if the game cannot end in a draw.
Worked example

A and B alternately throw a die until one rolls a six. A starts. Find each player’s probability of winning.

  1. A wins on their first, third, fifth, and subsequent odd-numbered turns.
    P(A wins)=16+(56)216+(56)416+⋯P(A\text{ wins})=\frac16+\left(\frac56\right)^2\frac16+\left(\frac56\right)^4\frac16+\cdots
  2. This is a geometric series with first term 16\frac16 and common ratio 2536\frac{25}{36}.
    P(A wins)=161−2536=611P(A\text{ wins})=\frac{\frac16}{1-\frac{25}{36}}=\frac6{11}
  3. Exactly one player eventually wins, so B’s probability is the complement.
    P(B wins)=1−611=511P(B\text{ wins})=1-\frac6{11}=\frac5{11}
Answer: P(A wins)=611P(A\text{ wins})=\frac6{11} and P(B wins)=511P(B\text{ wins})=\frac5{11}.
⚠ Common mistake

A’s later winning chances require failures on both players’ earlier turns, so the common ratio is (56)2(\frac56)^2, not 56\frac56.

Bayes’ Theorem: Machine Setup

Definition

Posterior probability of a setup — The probability that a machine setup is correct after observing its output is found by weighting each setup’s likelihood by its prior probability.

Model the production information

  • Prior probabilities: Correct setup B1B_1 has probability 0.80.8; incorrect setup B2B_2 has probability 0.20.2.
  • Likelihoods: Each item is acceptable with probability 0.90.9 under B1B_1 and 0.40.4 under B2B_2; for two independent items, the likelihoods are 0.920.9^2 and 0.420.4^2.
Probability of a correct setup after two acceptable items
P(B1∣A)=P(B1)(0.9)2P(B1)(0.9)2+P(B2)(0.4)2P(B_1\mid A)=\frac{P(B_1)(0.9)^2}{P(B_1)(0.9)^2+P(B_2)(0.4)^2}

Use independent-item probabilities to find the likelihood of the observed pair under each setup.

Using production evidence

  1. Assign prior probabilities to the possible setup states.
  2. Find the likelihood of the observed output under each state, using independence when appropriate.
  3. Apply Bayes’ theorem and simplify the posterior probability.
Worked example

A machine is correctly set up with probability 0.80.8. It produces acceptable items with probability 0.90.9 when correctly set up and 0.40.4 otherwise. Given that two items are acceptable, find the probability the setup is correct.

  1. Let AA be the event that both produced items are acceptable; calculate the joint weights for the two setups.
    P(B1)P(A∣B1)=0.8(0.9)2=0.648,P(B2)P(A∣B2)=0.2(0.4)2=0.032P(B_1)P(A\mid B_1)=0.8(0.9)^2=0.648,\quad P(B_2)P(A\mid B_2)=0.2(0.4)^2=0.032
  2. Divide the correct-setup weight by the total probability of the observed output.
    P(B1∣A)=0.6480.648+0.032=8185≈0.953P(B_1\mid A)=\frac{0.648}{0.648+0.032}=\frac{81}{85}\approx0.953
Answer: 8185≈0.953\frac{81}{85}\approx0.953, or about 0.950.95.
⚠ Common mistake

For two acceptable items, use (0.9)2(0.9)^2 and (0.4)2(0.4)^2 as the likelihoods; do not use the single-item probabilities alone.

Conditional Probability: Chapter Toolkit

Conditional probability rules

For P(F)>0P(F)>0, P(E∣F)=P(E∩F)P(F)P(E\mid F)=\frac{P(E\cap F)}{P(F)}. Also, 0≤P(E∣F)≤10\le P(E\mid F)\le1, P(E′∣F)=1−P(E∣F)P(E'\mid F)=1-P(E\mid F), and P((E∪F)∣G)=P(E∣G)+P(F∣G)−P((E∩F)∣G)P((E\cup F)\mid G)=P(E\mid G)+P(F\mid G)-P((E\cap F)\mid G).

Multiplication and independence

For nonzero conditioning probabilities, P(E∩F)=P(E)P(F∣E)=P(F)P(E∣F)P(E\cap F)=P(E)P(F\mid E)=P(F)P(E\mid F). Events EE and FF are independent when P(E∩F)=P(E)P(F)P(E\cap F)=P(E)P(F); equivalently, P(E∣F)=P(E)P(E\mid F)=P(E) and P(F∣E)=P(F)P(F\mid E)=P(F) whenever the conditional probabilities are defined.

Practice

Using the four-box data, find the probability that the drawn ball is black.

Hint: Use the total probability theorem with the four equally likely box choices.
Key step: Add each box-selection probability multiplied by its conditional black-ball probability.
Solution: P(A)=14(16+14+17+413)=9474368P(A)=\frac14\left(\frac16+\frac14+\frac17+\frac4{13}\right)=\frac{947}{4368}.
Practice

In the alternating die game, A starts and each player wins by rolling a six. Find the probability that B wins.

Hint: Use the complement of A’s winning probability.
Key step: First find A’s probability from the geometric series, then subtract it from 11.
Solution: P(B wins)=1−611=511P(B\text{ wins})=1-\frac6{11}=\frac5{11}.
Practice

A machine is correctly set up with probability 0.80.8 and incorrectly set up with probability 0.20.2. If two acceptable items are produced, find the posterior probability that the setup was incorrect.

Hint: Use the same two likelihoods as in the correct-setup calculation.
Key step: Divide the incorrect-setup weight 0.2(0.4)20.2(0.4)^2 by the sum of both setup weights.
Solution: P(B2∣A)=0.2(0.4)20.8(0.9)2+0.2(0.4)2=485≈0.047P(B_2\mid A)=\frac{0.2(0.4)^2}{0.8(0.9)^2+0.2(0.4)^2}=\frac4{85}\approx0.047.

Quick recap

  • A partition consists of pairwise disjoint events whose union is the sample space.
  • The total probability theorem adds the weighted likelihoods across a partition; Bayes’ theorem reverses the conditional probability.
  • Given that the drawn ball is black, the probability it came from box III is 156947≈0.165\frac{156}{947}\approx0.165.
  • For alternating die throws with A starting, the winning probabilities are 611\frac6{11} for A and 511\frac5{11} for B.
  • Given two acceptable items, the probability the machine was correctly set up is 8185≈0.953\frac{81}{85}\approx0.953.
  • Conditional probability, the multiplication rule, and independence connect joint and conditional events.

Formula sheet

Conditional probability
P(E∣F)=P(E∩F)P(F),P(F)>0P(E\mid F)=\frac{P(E\cap F)}{P(F)},\quad P(F)>0
Conditional probability
P(A∣B)=P(A∩B)P(B),P(B)>0P(A|B)=\frac{P(A\cap B)}{P(B)},\quad P(B)>0
Conditional probability for equally likely outcomes
P(A∣B)=n(A∩B)n(B)P(A|B)=\frac{n(A\cap B)}{n(B)}
Three or more events
P(E∩F∩G)=P(E)P(F∣E)P(G∣E∩F)P(E \cap F \cap G)=P(E)P(F \mid E)P(G \mid E \cap F)
Extension to n events
P(⋂i=1nEi)=P(E1)∏i=2nP(Ei∣⋂j=1i−1Ej)P\left(\bigcap_{i=1}^{n} E_i\right)=P(E_1)\prod_{i=2}^{n}P\left(E_i \mid \bigcap_{j=1}^{i-1}E_j\right)
Independence test
P(E∩F)=P(E)P(F)P(E \cap F)=P(E)P(F)
At least one of two independent events
P(A∪B)=1−P(A′)P(B′)=P(A)+P(B)−P(A)P(B)P(A \cup B)=1-P(A')P(B')=P(A)+P(B)-P(A)P(B)
Total probability
P(A)=∑j=1nP(Ej)P(A∣Ej)P(A)=\sum_{j=1}^{n}P(E_j)P(A\mid E_j)
Posterior probability of box III
P(E3∣A)=P(E3)P(A∣E3)∑i=14P(Ei)P(A∣Ei)P(E_3\mid A)=\frac{P(E_3)P(A\mid E_3)}{\sum_{i=1}^{4}P(E_i)P(A\mid E_i)}
Winning probabilities when A starts
P(S)=16,P(F)=56,P(A wins)=161−(56)2=611,P(B wins)=511P(S)=\frac16,\quad P(F)=\frac56,\quad P(A\text{ wins})=\frac{\frac16}{1-(\frac56)^2}=\frac6{11},\quad P(B\text{ wins})=\frac5{11}
Probability of a correct setup after two acceptable items
P(B1∣A)=P(B1)(0.9)2P(B1)(0.9)2+P(B2)(0.4)2P(B_1\mid A)=\frac{P(B_1)(0.9)^2}{P(B_1)(0.9)^2+P(B_2)(0.4)^2}

Mathematics — ALL CHAPTERS

1

Relations And Functions

View

2

Inverse Trigonometric Functions

View

3

Matrices

View

4

DETERMINANTS 

View

5

Continuity And Differentiability

View

6

Application Of Derivatives

View

7

Integrals

View

8

APPLICATION OF INTEGRALS

View

9

DIFFERENTIAL EQUATIONS

View

10

VECTOR ALGEBRA

View

11

Three Dimensional Geometry

View

12

Linear Programming

View

13

Probability

View

14

Proofs In Mathematics

View

15

Mathematical Modelling

View