3 ms·
Why is frequentism bad because it only gives certainty for infinite samples, but complexity is good despite being non-computable? It's two sides of the same co
by alisonkisk 6y ago
Why is frequentism bad because it only gives certainty for infinite samples, but complexity is good despite being non-computable?
It's two sides of the same coin -- computable uncertainty va non-computable certainty.
- EbTech 6y agoI don't think frequentism is "bad"; just insufficient as a gold standard interpretation of probabilistic claims. I liked an analogy from the reference by Rathmanner & Hutter: the most "correct" chess-playing program involves a complete search along the tree of possible games. In practice, we try to approximate this ideal. In the case of Kolmogorov complexity, a reasonable takeaway might be to use the shortest program that we're able to find, even if it's not the shortest overall.