Incentive Compatible

In this post I go into details of how statistical learning frames the relationship between out-of-sample and within-sample fit. I also connect it with the familiar idea of degrees of freedom in linear regression, which every student of econometrics encounters.

I was in a seminar a few years ago and the presenter, a fellow PhD student, was arguing in favor of his/her complex model by telling us the Akaike criterion for it was lower (hence better) than for the simple model.

The policy advice I gave made perfect economic sense. It just didn’t work out because of the political situation, and I’m not responsible for that

Every once in a while, I’ll import an R or Python library and I’ll get an error telling me that it can’t find some scary fortran library. This happened in R: