STAT 201B Lecture Notes - Lecture 29: Bayes Estimator

• School: Berkeley
• Department: Statistics
• Course: STAT 201B
• Professor: Elizabeth Purdom

Document Summary

E is bayes rule rcf , e) = int ref, e ) find a mini mes posterior risk. Rato , t ) ( ibuayay. sn?sgy average roo . Bayes rule " is unique admissible contradiction ef is not. 7- e rco , e ) e rca , et ) If not unique bayes rule , under conditions admissible. I = f- hii) under sq error loss bayes rules b- n bayes estimator i. Maxi max rule maximum risk estimator f minimax estimator rule sup rco , ekin f shop rco . Rto , c) = in tf -y if ett sup rca , ec) -_ a c-- i ce 5 ) rco , ea ) - E constant risk and e admissible then e minimax. Rco , e) = cto proof contradiction if e not minimax. Rio , e) e shot rca , e) a sup rco .