3 ms·
I haven't dug deeper, but the claim that the von Neumann approach is optimal does not seem intuitively correct. It seems like you could squeeze a tiny bit more
by andrewla 2y ago
I haven't dug deeper, but the claim that the von Neumann approach is optimal does not seem intuitively correct. It seems like you could squeeze a tiny bit more entropy from it -- basically, if you reject two pairs in a row, the nature of that rejection tells you something.
HT xx -> H
TH xx -> T
HH TT -> H
TT HH -> T
- nimish 2y agoYou're correct. There are more sophisticated extractors like Elias' and Peres' that do better, asymptotically achieving the upper bound given infinite data. See https://peteroupc.github.io/randextract.html https://peteroupc.github.io/randextract.html
- Maro 2y agoOP here. Thanks for that, that's obviously better in terms of bitrate, I haven't thought of that. Any two bit sequences with equal probability can be used to emit either a 0 and 1..