Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I don't think that's right. If you have enough data points, your prior gradually gets more relevant. And Bayesian statistics has the concept of an "ignorance prior", which mathematically represents the position where all possibilities are equally likely. Bayesian statistics also offers the possibility of adding more data after running your experiment, and computing a new answer in a consistent way. Whereas doing this with frequentist statistics completely is completely invalid.


"Ignorance priors" are not the common use case for Bayesian in general, at least in my experience.

Furthermore, who said adding more data is not allowed? Of course it is, and I agree fully, but do struggle to see how that contradicts what was said.

Think about a Bayesian spam filter, you keep training it with more "priors", over time and it gets better.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: