Hacker News new | ask | show | jobs
by svalorzen 1054 days ago
Also, in case anyone is interested, the uninformative Jeffreys prior for this in Bayesian statistics (meaning it does not assume anything and is invariant to certain transformations of the inputs) is Beta(0.5, 0.5). Thus the initial guess is 0.5, and it evolves from there from the data.
1 comments

Isn't 0.5 an absurd guess for the probability of a new restaurant being exceptionally good?