Course contentsShow
← Mathematics
Lesson 2817 of 3,12155. Probability — FoundationsFree lesson

Prior and Posterior Probabilities

Understanding how prior beliefs are updated to posterior probabilities when new evidence is observed.

Prior and Posterior Probabilities

What You'll Learn

You'll learn how prior beliefs are systematically updated to posterior probabilities when new evidence arrives—the heart of Bayesian reasoning.

The Core Idea

When you start with an initial belief about which hypothesis is true (your prior probability), and then observe new evidence, you can use Bayes' Theorem to calculate an updated, improved belief (your posterior probability).

Prior probability = P(H) — your initial assessment before seeing the evidence Posterior probability = P(H|E) — your updated assessment after observing evidence E

Think of it like a doctor's diagnosis: initially, the doctor has a baseline probability that you have a certain disease based on population statistics (prior). After seeing your test results (evidence), the doctor updates this to a more personalized probability (posterior).

The Update Process

Bayes' Theorem formalizes this update:

P(H|E) = P(E|H) · P(H) / P(E)
  • P(H): Prior — what we believed before
  • P(E|H): Likelihood — how likely this evidence is if hypothesis H is true
  • P(E): Normalizing constant — total probability of seeing this evidence
  • P(H|E): Posterior — our updated belief after seeing evidence E

The posterior incorporates both your prior knowledge and the new data. If the evidence is very consistent with hypothesis H, the posterior will be larger than the prior. If not, it will be smaller.

Example Flow

You estimate a 30% chance of rain today (prior). You step outside and see dark clouds (evidence). Given how often dark clouds lead to rain, you update your belief to 70% (posterior). Tomorrow, that 70% becomes your new prior for further updates.

Key Takeaway: Prior probabilities represent initial beliefs before evidence; posterior probabilities are the rational updates after incorporating new evidence via Bayes' Theorem—each posterior can become tomorrow's prior in sequential learning.