Psy, the “worst case expected value” is a bound directly comparable to the “worst case value” because the actual trajectory of expert weights is the same between the two algorithms.
Pearson, as an infinite set atheist I see no reason to drag infinities into this case, and I don’t recall any of the proofs talked about them either.
Psy, the “worst case expected value” is a bound directly comparable to the “worst case value” because the actual trajectory of expert weights is the same between the two algorithms.
Pearson, as an infinite set atheist I see no reason to drag infinities into this case, and I don’t recall any of the proofs talked about them either.