Why do we take `(Bias) ^2` in total error in a model?

Asked Jan 13 '19 at 12:58

Active Jan 15 '19 at 14:05

Viewed 31 times

I was recently studying some book and few blogs and come to note that :

Total error = Bias^2 +Variance + irreducible error

Also, I know that these are the errors but I have problem understanding, How do we quantify these errors? What I mean by this is : Give few data points and the predicted and the true value. How do we measure Bias, variance and irreducible error?

edited Jan 15 '19 at 14:05

kjetil b halvorsen

63,378
26
142
467

asked Jan 13 '19 at 12:58

Akash Dubey

@Xi'an How is it related to pythagoras theorem ? Can you please explain ? – Akash Dubey Jan 13 '19 at 15:57
Can you provide a link to the blogs or a reference to the book? – Mark White Jan 13 '19 at 17:27
1

1) from dimensional analysis alone it is clear that bias must be squared ... 2) search this site for error decomposition (or bias-variance decomposition) must be answered before. – kjetil b halvorsen Jan 15 '19 at 14:04

Why do we take `(Bias) ^2` in total error in a model?

0 Answers0