I wish there were more academics from nonparametric statistics trying to formally relate cross-validation to things like the jackknife and exact permutation tests.
From the perspective of a non-developer I thought this was pretty good.
I could have done with a bit more information and references, and less arguments with "past-me" though.
HN has an almost hostile reaction to anyone who says using machine learning in practice is easy. Maybe it is just a question of different definitions.
Of course developing new algorithms and coding everything from scratch is difficult but using machine learning for predictive analytics does not have to be that hard and can create a massive amount of business value.
Thank you for sharing. I don't have any experience with predictive analytics. I found this article approachable and it sparked my interest in the subject.
5 comments
[ 2.7 ms ] story [ 21.5 ms ] threadI wish there were more academics from nonparametric statistics trying to formally relate cross-validation to things like the jackknife and exact permutation tests.
Or whatever those crazy kids do these days.
HN has an almost hostile reaction to anyone who says using machine learning in practice is easy. Maybe it is just a question of different definitions.
Of course developing new algorithms and coding everything from scratch is difficult but using machine learning for predictive analytics does not have to be that hard and can create a massive amount of business value.