Course contentsShow
← Mathematics
Lesson 2816 of 3,12155. Probability — FoundationsFree lesson

Bayes' Theorem with Multiple Hypotheses

Extending Bayes' Theorem to choose among several competing hypotheses using the law of total probability.

Bayes' Theorem with Multiple Hypotheses

What You'll Learn

You'll learn how to use Bayes' Theorem to update beliefs across several competing explanations when new evidence arrives.

The Core Idea

So far, you've used Bayes' Theorem to flip conditional probabilities when you have just one hypothesis (like "has the disease") versus its complement ("doesn't have the disease"). But real life often presents multiple competing explanations—imagine three different factories producing the same item, or several possible causes for a medical symptom.

Bayes' Theorem with multiple hypotheses lets you calculate the probability of each specific hypothesis given observed evidence, by combining Bayes' formula with the law of total probability.

The Setup

Suppose you have hypotheses H₁, H₂, ..., Hₙ that partition the sample space (mutually exclusive and exhaustive). You observe evidence E. For each hypothesis Hᵢ:

P(Hᵢ | E) = P(E | Hᵢ) · P(Hᵢ) / P(E)

where the denominator expands using the law of total probability:

P(E) = P(E | H₁)·P(H₁) + P(E | H₂)·P(H₂) + ... + P(E | Hₙ)·P(Hₙ)

Practical Meaning

Think of it like this: You start with prior beliefs about how likely each hypothesis is (the P(Hᵢ) values). When evidence E appears, you weight each hypothesis by how well it "predicts" that evidence (the P(E | Hᵢ) values). The denominator normalizes everything so all posterior probabilities sum to 1.

Each hypothesis gets updated proportionally to both its prior plausibility and how compatible it is with the new data.

Key Takeaway: Bayes' Theorem extends naturally to multiple competing hypotheses by using the law of total probability in the denominator, giving you updated probabilities for every hypothesis simultaneously after observing evidence.