I graduated with a Ph.D. in Statistics, in Fall Quarter, 2000 from the Department of Statistics at the University of Washington , Seattle, under the supervision of Professor Jon A. Wellner . Prior to graduate study at the University of Washington, I was a student at the Indian Statistical Institute , where I received my Bachelors (B.Stat.) and Masters (M.Stat.) in Statistics with specialization in Mathematical Statistics and Probability. For more detailed information about my academic and professional background, see my CV. Apart from Statistics I take a general interest in the arts and humanities and especially in poetry, classical music, philosophy and drama.
Research:My dissertation work at UW deals with likelihood ratio inference in a class of regular and non-regular problems. The part dealing with regular problems addresses likelihood ratio inference in regular semiparametric models where interest focuses on the estimation of some finite dimensional function of the indexing parameter. The other part, dealing with non-regular models, concerns likelihood ratio inference in the setting of non-parametric estimation of a monotone function. The likelihood ratio statistic in this scenario is no longer asymptotically chi-squared but does have a universal limiting distribution. For more details on this, as well as my work as a Research Assistant at the department and the kind of research I would like to do in future, see my CV and the following description of my research.
Slides for my talks on likelihood ratio inference for monotone functions.
Currently, we are working on characterizing the power behavior of the likelihood ratio statistic in certain non-regular situations, both under fixed and local(contiguous)alternatives. A related issue of interest is the comparison of the likelihood ratio statistic with some other statistics suggested by Groeneboom, which are based on L^2 distances between unconstrained and constrained MLE's. Slides and plots for this current work presented at the recent INVERSE PROBLEMS conference in Eindhoven, The Netherlands are available by clicking on this link.
Talk for Empirical Processes Working Group:I will follow my notes on Profile Likelihood for my talk for the working group. Other references are the 1997 Annals Paper on Semiparametric Likelihood Ratio Inference by Murphy and Van der Vaart, the 1999 JASA paper "On Profile Likelihood" by the same authors and Chapter 2 of my thesis that's downloadable off this website, or a more condensed version of it which appears right below (the first link) under "Papers and Technical Reports". The two Murphy and Van der Vaart papers can be downloaded off Susan Murphy's homepage.
Papers and Technical Reports:Banerjee, M. (2000): Semiparametric Likelihood Ratio Inference Revisited. UW Department of Statistics Technical Report No. 384 .
Banerjee, M. and Wellner J.A. (2000): Likelihood Ratio Tests for Monotone Functions. UW Department of Statistics Technical Report No. 377. Submitted to Annals of Statistics.
Banerjee, M. and Richardson, T. (2000): A Note on "On a Dualization of Graphical Gaussian Models". UW Department of Statistics Technical Report No. 376.
Richardson, T., Bailer, H. and Banerjee, M. (1999): Specification searches using MAG models. Proceedings, ISI Conference, Helsinki, 1999.
Richardson, T., Bailer, H. and Banerjee, M. (1999): Tractable structure search in the presence of latent variables . In Proceedings of Artificial Intelligence and Statistics '99 (D. Heckerman and J. Whittaker, eds.), Morgan Kaufmann, San Francisco, CA, pp.142-151.
TEACHING: Statistics 220 in Winter Quarter and Spring Quarter, 2001.
Contact Information:| Address: | University of Washington Department of Statistics Box 354322 Seattle, WA 98195-4322 |
|---|---|
| Phone: | (206) 543-8471 |
| Email: | mouli@stat.washington.edu |