Bayesian magic is taking the same information, performing the same math, and making widely different statistical inferences! Don't like talking in classical terms, stick a flat prior on it!
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I agree
(unrealistic) infinite flat-prior is a reason of overfitting!
But if we assume finite prior range R of parameter, we can marginalize this unknown R and avoid overfitting
(that’s how I could derive MAP estimate of PCA’s dimension for the 1st time)https://twitter.com/viettran86/status/1045482416914087937?s=20 …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.