Lin Hsin Hsin
Artificial Intelligence Center
Bayesian Probability Distribution
P(B|A) P(A)
P (A|B) = __________________________
P(B)
Objective vs Subjective
Objectivists
Probability as an extension of logic
Probability quantifies reasonable expectation
According to Bayesian rules: Everyone should share
Subjectivists
Probability corresponds to a personal belief
rationality & coherence allow for
substantial variations with constraints
constraints are justified by:
Dutch book argument
Decision Theory
de Finetti Theorem
δ
Objective vs subjective variants of Bayesian Probability
differ mainly in their interpretation
& construction in the prior probability
Prior state of knowledge defines
(unique) prior probability distribution
📊
Applications
Supervised Learning
Unsupervised Learning