Law of Total Probability

Problem 2.4 • Breaking down P(B) through a partition of the sample space

P(B) = Σᵢ P(B|Aᵢ) · P(Aᵢ)

Sum over all mutually exclusive, exhaustive events Aᵢ

🌳 Probability Tree

📊 Partition of Sample Space

A₁
P(A₁) = 0.30
P(B|A₁) = 0.80
A₂
P(A₂) = 0.50
P(B|A₂) = 0.40
A₃
P(A₃) = 0.20
P(B|A₃) = 0.10

Adjust Prior Probabilities

P(A₁)
P(A₂)
P(A₃)

📐 Calculation

P(B|A₁)·P(A₁) 0.240
P(B|A₂)·P(A₂) 0.200
P(B|A₃)·P(A₃) 0.020
P(B) = Sum 0.460
Total Probability P(B)
0.460

💡 Key Insight

The law of total probability lets us calculate P(B) by considering how likely B is under each possible "cause" Aᵢ, weighted by how likely each cause is. This is the foundation for Bayes' theorem!