Re:Machine Learning Crash Course: The Bias-Variance Dilemma
by xapata on Wednesday July 26, 2017 @03:35AM
(#14854275)
Ultimately, everything is a probabilistic model. It'd be ridiculously impractical to try to incorporate all of one's knowledge into an analytic model. Instead, we generalize as appropriate and encapsulate the rest as a random variable, sometimes called "error". Plus, God plays dice and all that.
[ Parent
| on HN ]
There are no comments on this comment yet.