STAT 201B Lecture Notes - Lecture 12: Asymptotic Distribution, Fisher Information, Vale Limited

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

Document Summary

Asymptotic normality me assumptions (cid:15482) nco , i ) smoothness. In co) = va #inco ) ) = nico ) Icf) = val# log f- cx ; ol ) = - eoc# log fix ; o ) ) = zoe # - d= - e logo -o -eogx] I - plug in into in lo) en ii -965 : observed fisher. Tinformation t curvature t variability f- inc e) = iced th jn ce) Ice) approx e for true o approx average for expectation if ic ) continuous. Ico : e tineo) still random quantity while th inco) exactly. I , # log fcxi ; o ) to. I fico - 07 (cid:15482) nco , o) Tn ce - o ) (cid:15482) nco , fil ) If en any other estimator tn ce - 07 -7n conca) T cx) -_ ex , - - - - xn ) flex :o) -