This Exam P sample reference tests Bayes' Theorem. The zero-count likelihoods are exp(-0.5) and exp(-1) for the two intersection classes. Bayes' theorem gives 0.4exp(-0.5)/[0.4exp(-0.5)+0.6exp(-1)]=0.5236, so choice D is correct.
These notes identify the calculation error associated with each wrong letter when that error is reproducible.
AThis is approximately the joint probability 0.4 exp(−0.5)=0.2426 and omits normalization by the total probability of zero events.
BThis repeats the prior probability 0.40 and ignores the evidence supplied by the observed zero count.
CThis is approximately the evidence probability 0.4 exp(−0.5)+0.6 exp(−1)=0.4633, not the posterior class probability.
EUsing 0.4 for both prior weights gives exp(−0.5)/[exp(−0.5)+exp(−1)]=0.6225 and erases the 60% prior weight of the higher-rate class.
Original practice · fully worked
Original variant: reconstructing a shipment prior
A shipment arrives from either North Dock or South Dock. A scan flags 20% of North Dock shipments and 50% of South Dock shipments. Among flagged shipments, the probability of North Dock origin is 1/3. Calculate the prior probability that a shipment comes from North Dock.
A 0.1667
B 0.3333
C 0.5000
D 0.5556
E 0.7143
Variant answer in brief
Writing the unknown prior as p and applying Bayes gives 0.2p/[0.2p+0.5(1-p)]=1/3. Solving yields p=5/9, so choice D is correct.
Setup
Setup
Let p be the unknown prior probability of North Dock origin and let F denote a scan flag.
Pr(N)=p,Pr(F∣N)=0.2,Pr(F∣Nc)=0.5
Model
Model
Express the supplied posterior probability in terms of the unknown prior.
0.2p+0.5(1−p)0.2p=31
Compute
Compute
Cross-multiply and solve the resulting linear equation.
The 3108-page Probability Proof Manual reorganizes 718 verified Exam P solutions by syllabus skill and adds formula proofs, error patterns, and original worked practice.