News for hackers. Stuff that matters.
Re:Machine Learning Crash Course: The Bias-Variance Dilemma by creeble on Monday July 17, 2017 @04:08PM (#14791901)
I also think that, even if all the information is efficiently baked into price, price isn't the predictor - the information doing the baking is. If you use the "predictant" as the predictor, you won't get very good results.

Not to say that people don't, but as mentioned, they don't last. How long they do last is an even more difficult prediction problem.

There are no comments on this comment yet.