BS2 Statistical Inference - Lecture 3

Sufficiency and unbiased estimation

This lecture is concerned with the interplay between unbiasedness, minimum variance, sufficiency and completeness. In particular we show that the likelihood function is minimal sufficient and that a MVUE is always a function of a minimal sufficient statistic. Rao-Blackwell's theorem is established and the important idea of Rao-Blackwellization is introduced.

Overheads

screen viewing

printing

Links

Next lecture

Previous lecture

Course overview


Last updated: Friday, 22 October 2004 11:39Steffen L. Lauritzen