nConditional or posterior
probabilities
ne.g., P(cavity | toothache) =
0.8
ni.e., given that toothache is
all I know
n
nNotation for conditional distributions:
nP(cavity | toothache) = 2-element vector of 2-element vectors
n
nIf we know more, e.g., cavity is also given, then we
have
nP(cavity
|
toothache, cavity) = 1
n
nNew evidence may be irrelevant, allowing simplification,
e.g.,
nP(cavity
|
toothache. sunny) = P(cavity | toothache) =
0.8
nThis kind of inference, sanctioned by domain knowledge,
is crucial