One hundred icons split into 80 cyan A's and 20 orange B's illustrate how Bayes' rule updates beliefs after evidence. Evidence E appears in 10% of A and 60% of B; highlighting those cases regroups 20 symbols whose proportions give posteriors P(A|E)=0.4 and P(B|E)=0.6. Each term of Bayes' formula—prior, likelihood, joint, and normalizer—is tied to a visual group. Changing the priors to 0.95/0.05 shows the posterior for B drop to 0.24, clarifying why likelihood and posterior differ. Useful for first-year AI and statistics students.
Narrated · 16:9 · Preview before teaching · automatic layout checks do not establish subject accuracy
# Bayes' rule changes the population we compare One hundred symbols represent two classes: 80 cyan A symbols and 20 orange B symbols. Evidence E occurs in 10 percent of class A and 60 percent of class B. Highlight 8 A symbols and 12 B symbols, then move these 20 symbols into a separate group representing the observed evidence. The proportions within this group become the posterior probabilities: P(A given E) = 0.4 and P(B given E) = 0.6. Build Bayes' rule from the colored groups, linking the prior, likelihood, joint probability, and normalizing denominator to their visual counterparts. Under equal error costs, choose B because its posterior is larger. Briefly change the priors to 0.95 and 0.05 while keeping the likelihoods fixed; the posterior for B becomes 0.24. Use a black background and reveal one calculation at a time. Useful for distinguishing likelihood from posterior probability. Audience: 一年级人工智能专业的学生