9 ms·
In my admittedly very cursory skim of the paper, I gave it the benefit of the doubt on this point, because this seemed convincing: "for the purpose of predicti
by voidmain 8y ago
In my admittedly very cursory skim of the paper, I gave it the benefit of the doubt on this point, because this seemed convincing:
"for the purpose of prediction, we are, of course, not so much interested in the probabilities issued by the measure functions, as in the conditional probabilities that give the corresponding predictors’ outputs."
Do you disagree?
(Also, my original summary is wrong at least in that I should have said limit computable where I said semi computable!)
- cousin_it 8y agoI guess the question is whether we want universality over all limit-computable predictors, or only those derived from lower-semicomputable semimeasures. Since the latter class contains its own universal element and also every computable predictor, maybe it's okay? But yeah, I can see how you can have a different takeaway. Sorry for being a bit harsh in the previous comment.
- voidmain 8y agoNo need to apologize, I thought your comment was perfectly civil. It can be hard to read tone on the internet!