Bayes
Comprehensive notes, formulas, and practice questions for Bayes.
Bayes
Bayes' Theorem
What you'll learn
- Bayes' theorem for reversing conditional probability: from P(B|A) to P(A|B).
- Combining law of total probability with Bayes for diagnostic and classification problems.
- To set up prior, likelihood, and posterior probabilities in medical testing and quality control.
- To interpret results when base rates are small — why a positive test may still mean low disease probability.
- JEE/NEET applications: factory defects, disease screening, and spam filtering intuition.
Key concepts
Level 1 — Foundations
Verbal: Bayes' theorem tells us how to update probability of a cause (hypothesis) after observing evidence. It connects "effect given cause" with "cause given effect."
Formula: P(A|B) = P(B|A)·P(A) / P(B), when P(B) > 0.
Expanded (partition B₁,…,Bₙ): P(Aᵢ|B) = P(B|Aᵢ)P(Aᵢ) / Σⱼ P(B|Aⱼ)P(Aⱼ)
Terminology:
| Term | Meaning |
|---|---|
| P(A) | Prior — before evidence |
| P(B | A) |
| P(A | B) |
Level 2 — JEE / NEET depth
Law of total probability (denominator setup): P(B) = Σ P(Aᵢ)P(B|Aᵢ) over partition {Aᵢ}.
Medical test classic: Disease prevalence P(D) small; test sensitivity P(+|D) high; false positive rate P(+|D′) matters. Bayes gives P(D|+) — often surprisingly low when prevalence is low.
Factory QC: Machines M₁, M₂ produce items; defect rates differ. Observing defect, Bayes finds which machine likely produced it.
Sequential updating: Posterior from one observation becomes prior for next — foundation of Bayesian inference.
JEE strategy:
- Identify hypothesis partition (causes).
- Write given P(Aᵢ) and P(B|Aᵢ).
- Compute denominator via total probability.
- Apply Bayes for requested P(Aᵢ|B).
Independence note: Bayes still applies when events are dependent; independence simplifies multiplication only.
Worked example
Medical screening problem
Disease prevalence 1%. Test: 99% sensitive P(+|D), 5% false positive P(+|D′). Find P(D|+).
Step 1 — P(D) = 0.01, P(D′) = 0.99.
Step 2 — P(+|D) = 0.99, P(+|D′) = 0.05.
Step 3 — P(+) = (0.99)(0.01) + (0.05)(0.99) = 0.0099 + 0.0495 = 0.0594.
Step 4 — P(D|+) = (0.99)(0.01)/0.0594 ≈ 0.167 ≈ 16.7%.
Step 5 — Despite positive test, only ~17% chance of disease due to low base rate.
Factory machine defect attribution
M₁ produces 60% items (2% defective); M₂ produces 40% (5% defective). Item is defective. P(from M₂)?
Step 1 — P(M₂|Def) = P(Def|M₂)P(M₂) / P(Def).
Step 2 — P(Def) = 0.02(0.6) + 0.05(0.4) = 0.012 + 0.02 = 0.032.
Step 3 — P(M₂|Def) = (0.05)(0.4)/0.032 = 0.02/0.032 = 0.625.
Step 4 — Higher per-item defect rate on M₂ makes it more likely source despite smaller share.
Common mistakes
| Mistake | Why it happens | Fix |
|---|---|---|
| Ignoring prior/base rate | Focusing only on sensitivity | Always include P(A) in numerator and total probability in denominator |
| Using P(A | B) = P(B | A) |
| Partition not exhaustive | Missing causes | Hypotheses must cover all possibilities and be mutually exclusive |
| Forgetting false positives in medical tests | Using sensitivity alone | Include P(+ |
Quick check
- State Bayes' theorem in words and symbols.
- If P(A)=0.3, P(B|A)=0.8, P(B|A′)=0.2, find P(A|B).
- Why can P(D|+) stay low with accurate test?
- Two urns: Urn I (3W,2B), Urn II (1W,4B). Random urn, draw white. P(Urn I)?
- Stretch: Extend Bayes to three-machine partition with unequal outputs.
NCERT Chapter 13 link: Bayes' theorem is Example 13.6 onwards — medical and factory problems dominate. Always define partition events (H₁, H₂, …) covering all cases with stated prior probabilities.
Exam connections: NEET-style questions emphasise low prevalence leading to surprising posterior — explain in words, not only numbers. JEE may embed Bayes inside longer probability chains — compute total probability denominator carefully with all partition terms.
Study strategy: Template: list causes, write P(cause), P(evidence|cause), multiply for joint, sum for P(evidence), divide for posterior. Table format prevents omission. Practice one medical and one urn problem until automatic.
Study workflow and exam preparation
When studying Bayes' Theorem within Probability, start by listing every formula and definition on one page without looking at the textbook. Compare your list to NCERT — missing items indicate gaps to fix immediately. Work through at least two NCERT Examples for this section with steps written in full; examiners award method marks even when arithmetic slips.
For board exams (CBSE), long answers benefit from a clear structure: definition → explanation → diagram or formula → example → brief conclusion. Underline key terms. For JEE Main and NEET, prioritise conceptual traps and quick calculation paths; timed mixed quizzes of 10 questions after revision simulate exam pressure.
Cross-topic link: Coordinate geometry and vectors often combine with matrices; calculus links to physics kinematics problems.
Spaced revision: Review this note at 1 day, 3 days, and 7 days after first study. Attempt the Quick check questions closed-book, then open the Practice tab for graded reinforcement. Maintain an error log — repeated mistake patterns reveal whether the issue is concept, formula recall, or careless reading.
Diagram and terminology drill: For Mathematics, redraw key figures from memory and define every labelled part in one sentence. Vocabulary precision prevents mark loss in descriptive answers — use NCERT terms exactly as printed in the textbook.
Revision tip: Link this topic to adjacent Class 12 chapters before attempting mixed practice.
Open the Practice tab for graded questions on Bayes' Theorem.
Key Takeaways (TL;DR)
- What you'll learn
- Key concepts
- Worked example
- Common mistakes
Master this topic with Drishti OS
Get unlimited mock tests, AI-powered mentorship, and complete video courses when you join.
Start Free Practice