Question:
Can we use lifting in probabilistic case?
nDo
probabilistic inference about “generic” objects x
nOnly make
distinctions that affect relevant probabilities
[Pfeffer
& K.]
Rest
of
model
student x
We can extend these ideas to
the probabilistic case, albeit in a more limited way than in theorem
proving.Rather than reasoning
separately for each student in our domain, we do probabilistic inference for
a “generic” student x, and then assume we have the right number of copies.