@StrongBad Are you referring to a reviewer in the sense of a peer-reviewed article? Or a reviewer commenting on a question or answer posted on CV.SE?

@usεr11852 I agree with your comment - It seems the question arises from a confusing difference in the terminology and precise definition of empirical vs percentile - I would recommend a thorough check by knowledgeable people and then editing the question to make the conclusions clear. Because as it reads now you would think percentile bootstrap is bad, whereas Efron (the inventor of bootstrap) dedicates an entire chapter of his new book explaining why it is great...

@MatthewDrury @StephanKolassa A detailed question about how to correctly use class weights with Random Forest. stats.stackexchange.com/questions/282947/…

I just saw that!

I've got some chores to do, but will try to work on an answer this weekend.

@Sycorax Thanks! I wanted to only paste a comment, so as not to steal @MatthewDrury's thunder, but then I kind of got carried away and realized that it turned into an answer, along these lines...

Sorry, Matt! I'll happily upvote your answer (given a strong prior that it will be of high quality)!

hi

I am looking for a good global optimization package for python? I know about docs.scipy.org/doc/scipy/reference/generated/… but I thought I read there better packages

any recommendations?

@Anush: try NLopt. More involved answers/questions should be directed to Computational Science.

@usεr11852 thanks.. I was wondering about hyperopt. Have you tried that?

Nope, sorry.