The probability function I chose meets the requirements in the paper; therefore it is a case that the theorem should apply to.
Trying to set p(n) to 1/U(n) * 1/2^n is clever; but it doesn’t work, because that probability distribution is not known to sum to 1. (It would sum to 1 if U(n) = 1 for all n.)
The probability function I chose meets the requirements in the paper; therefore it is a case that the theorem should apply to.
Trying to set p(n) to 1/U(n) * 1/2^n is clever; but it doesn’t work, because that probability distribution is not known to sum to 1. (It would sum to 1 if U(n) = 1 for all n.)