A Maximum Likelihood Oddity

The next talk in the La Trobe University Department of Mathematics and Statistics Colloquium Series will be given by Dr Hien Nguyen.

Seminar Abstract:

Ever since its popularisation by R.A. Fisher in the early nineteen hundreds, the method of maximum likelihood estimation has grown to become among the most useful and most used point estimation techniques in statistics and allied fields. Under generous regularity conditions, the maximum likelihood estimator for any particular probability model is easy to obtain and has numerous desirable properties such as consistency and efficiency. Unfortunately, there are some very simple probability models for which the maximum likelihood estimator is not so well behaved and where it is not so simple to state. One such model is the single parameter triangular distribution over the unit interval. We shall discuss some curiosities that arise from the maximum likelihood estimation problem for data arising from such distributions and present some recent results and an open question that is left unknown under such scenarios.

Speaker Bio:

Host University: La Trobe University

Seminar Convener: Dr Hien Nguyen

How to participate in this seminar

1. Book your University’s Access Grid Room, or a university or APAC etc. Access Grid Room that you are able to use, and

2. Send an email to the seminar convener at the University of Newcastle (Scott Lindstrom) to advise that you will be attending the seminar.