BS2 Statistical Inference - Lecture 8
Asymptotic properties and computation of the MLE
This lecture goes somewhat deeper into asymptotic results for maximum likelihood estimates. We finalise the proof of asymptotical normality and efficiency in the general case, assuming consistency of the MLE and sketch proofs of Cramér and Wald.
Alternative computations of the variance of the MLE using for example the observed Fisher information is briefly mentioned and the Newton-Raphson method for computation of the MLE is described.
Overheads
Links
Last updated: Monday, 08 November 2004 19:25. Steffen L. Lauritzen