bayes network: inference Flashcards
(16 cards)
what is diagnosis in bayes network
P(Cause | Effect)
what is Prediction in bayes network
P(Effect | Cause)
applications of inference in bayes network
classification maxclassP(class | data)
decision making: P(effect | Cause ) x Utility(Effect | Cause)
what is marginal distribution
a distribution formed by calculating the subset of a larger probability
what are the 3 categories of variables
evidence variables
query variables
non - evidence variables
what is a evidence variable
known variables
what is a query variable
wanted variables
what is a non evidence variable
neither known nor wanted , but must deal with
what are the complete set of variables in bayes network
E U X U Y
basically all 3 variables together
what is exact inference equation
P(X | E) = α·P(X ,E ) = α·P(X ,E ) = α· sumof Y
P(X ,E ,Y )
what does α in exact inference
α = 1 / P (E)
how do we find discrete random variables in marginalisation
we sum over unwanted variables
how do we find continuous random variables in marginalisation
integrate over unwanted variables
what is exact inference in general
Bayesian inference algorithms that calculate the exact
value of probability P(X |E )
inference by enumeration
We infer a posterior probability by marginalization of the joint distribution.
That is, we compute the exact value of probability P(X |E ) using the
equation
computational complexity
The number of terms in the sum is exponential in the number of
non-evidence random variables: The complexity is O(nmn), where:
▶ n is the number of non-evidence variables.
▶ m is the number of values each variable can take.
