Lecture 12: Monte Carlo Integration (25)
ziyaointl

Unbiased estimators are preferred because we can simply take its value as an estimate for the desired integral without having to apply any corrections.

JiaweiChenKodomo

Also, according to the law of large numbers, we are confident that the estimator will converge to the correct answer as we sample it more and more. This "certainty" is desirable.

marianavazquezr

I am trying to think of an example where biased estimators would be preferred over unbiased estimators but I cant really think of any. Does anyone have any examples?

caokevinc

In lecture Ren briefly mentions that there are biased estimators that are very computationally quick, which could be used in situations where speed matters more than accuracy.

YoungNathan

Are there any examples where a biased estimator may be used to produce things like stylistic effects rather than realism?

You must be enrolled in the course to comment