Probability

NCERT Class 12 probability — conditional probability, Bayes' theorem, random variables, and probability distributions including the binomial. From predicting election outcomes to assessing medical test accuracy to designing fair games, probability is the mathematical language of uncertainty. This chapter extends the foundational concepts from Class 11—sample spaces, events, and basic probability—to sophisticated tools: conditional probability (probability when you have partial information), independent events (when one event doesn't influence another), and Bayes' theorem (updating beliefs with evidence). These concepts underpin statistics, machine learning, and decision-making under uncertainty.

Recalling Foundations from Class 11

Class 11 introduced:

• Sample space S: all possible outcomes of an experiment
• Event E: a subset of S (collection of outcomes of interest)
• Classical probability: P(E) = |E|/|S| (number of favorable outcomes / total outcomes)
• Empirical probability: P(E) = frequency of E / total trials

Axioms of probability (Kolmogorov):

• 0 ≤ P(E) ≤ 1 for any event E
• P(S) = 1 (certainty for entire sample space)
• For mutually exclusive events: P(A ∪ B) = P(A) + P(B)

This chapter builds on these foundations to handle complex, real-world probability questions.

Conditional Probability

Conditional probability P(A|B) is the probability of event A given that event B has occurred:

P(A|B) = P(A ∩ B) / P(B)     (assuming P(B) > 0)

Intuition: We restrict our sample space to outcomes in B. Among those outcomes, what fraction are also in A?

Example: Drawing cards without replacement from a deck. P(2nd card is ace | 1st card was ace) = 3/51 (only 3 aces left among 51 cards), not 4/52.

Key formula: P(A ∩ B) = P(A|B) · P(B) = P(B|A) · P(A) (multiplication rule)

This allows us to find the probability both events occur if we know conditional and marginal probabilities.

Multiplication Rule and Tree Diagrams

For sequence of events:

P(A₁ ∩ A₂ ∩ ... ∩ Aₙ) = P(A₁) · P(A₂|A₁) · P(A₃|A₁,A₂) · ... · P(Aₙ|A₁,...,Aₙ₋₁)

Tree diagrams: Visualize sequential experiments with branches representing outcomes. Probability of a path is the product of conditional probabilities along the path.

Example: Two-stage experiment: Urn A (3 red, 2 blue balls), draw one, transfer to Urn B (2 red, 3 blue), draw one from B.

• P(both red) = P(red from A) · P(red from B | red from A) = (3/5) · (3/5) = 9/25
• P(first red, second blue) = (3/5) · (2/5) = 6/25

Independent Events

Events A and B are independent if the occurrence of one doesn't affect the probability of the other:

P(A|B) = P(A)  or equivalently  P(A ∩ B) = P(A) · P(B)

Example: Flipping a coin twice. First flip being heads doesn't change P(heads on 2nd flip) = 1/2. These are independent.

Contrast: Drawing cards without replacement. Probability of 2nd card depends on 1st card. Events are dependent.

For n independent events:

P(A₁ ∩ A₂ ∩ ... ∩ Aₙ) = P(A₁) · P(A₂) · ... · P(Aₙ)

Bayes' Theorem

Bayes' theorem updates probabilities when new evidence arrives:

P(A|B) = P(B|A) · P(A) / P(B)

Interpretation:

• P(A|B): posterior probability (probability of A after observing B)
• P(B|A): likelihood (probability of observing B if A were true)
• P(A): prior probability (initial probability of A before evidence)
• P(B): marginal probability (total probability of observing B)

Example: Medical diagnosis. A disease affects 1% of population. A test is 99% accurate (detects disease if present, correctly identifies absence if not present). You test positive. What's the probability you have the disease?

= 0.99 · 0.01 + 0.01 · 0.99 = 0.0198

• P(disease) = 0.01 (prior)
• P(positive | disease) = 0.99 (likelihood)
• P(positive | no disease) = 0.01 (false positive rate)
• P(positive) = P(positive | disease) · P(disease) + P(positive | no disease) · P(no disease)

Using Bayes:

P(disease | positive) = (0.99 · 0.01) / 0.0198 ≈ 0.505

Surprisingly, you have only about 50% chance of having the disease despite a positive test! This reveals the importance of prior probability.

Total Probability and Partition Formula

If events B₁, B₂, ..., Bₙ partition the sample space (mutually exclusive and exhaustive):

P(A) = Σ P(A|Bᵢ) · P(Bᵢ)

This decomposes a complex probability into simpler conditional probabilities.

Example: A company has three factories producing 30%, 50%, and 20% of items. Defect rates are 2%, 1%, and 3% respectively. What's probability a random item is defective?

P(defective) = P(defective|F₁)·P(F₁) + P(defective|F₂)·P(F₂) + P(defective|F₃)·P(F₃)
             = 0.02·0.30 + 0.01·0.50 + 0.03·0.20
             = 0.006 + 0.005 + 0.006
             = 0.017

Random Variables and Probability Distributions

A random variable X is a function assigning numerical values to outcomes.

Example: Flip coin twice; X = number of heads. X can be 0, 1, or 2.

A probability distribution lists all possible values and their probabilities:

X        0    1    2
P(X)   1/4  1/2  1/4

Expected value (mean): E[X] = Σ x · P(X = x) For the coin example: E[X] = 0·(1/4) + 1·(1/2) + 2·(1/4) = 1

Variance and standard deviation measure spread of distribution.

Connections to Other Topics

• Class 11 Basic Probability: Foundation of sample spaces, events, classical probability
• chapter-09-differential-equations: Stochastic processes and probability distributions
• Statistics (beyond Class 12): Hypothesis testing, confidence intervals, regression
• Machine learning: Bayes' theorem underpins many algorithms
• Decision theory: Expected value guides choices under uncertainty

Socratic Questions

1. Explain the difference between P(A|B) and P(B|A) using a concrete example. How does conditioning on different events change the probabilities, and why are these generally not equal?

1. Bayes' theorem says P(A|B) = P(B|A) · P(A) / P(B). In the medical diagnosis example, why does the prior probability P(A) matter so much? What would change if the disease were rarer (say 0.1% of population)?

1. Two events are independent if P(A|B) = P(A). Explain why independence means that learning B occurred gives you no information about whether A occurred. Can you think of events that seem unrelated but might be dependent?

1. For the multiplication rule P(A ∩ B) = P(A|B) · P(B), explain why you multiply conditional probability by P(B). How would this change if you knew P(B|A) instead of P(A|B)?

1. In the total probability formula, why must the events B₁, B₂, ..., Bₙ partition the sample space (be mutually exclusive and exhaustive)? What goes wrong if you apply the formula to events that overlap?