McMASTER UNIVERSITY STATISTICS SEMINAR

Week of February 9 - 13, 1998

SPEAKER:	Dr. Reg Kulperger Department of Statistical and Actuarial Sciences, University of Western Ontario
TITLE:	"Empirical Processes and Tests of Randomness"
DAY:	Wednesday, February 11, 1998
TIME:	3:30 p.m. [Coffee in BSB-202 at 3:00 p.m.]
PLACE:	BSB-108

SUMMARY

This talk will consider some methods to test if an observed sequence X₁, X₂, ... , X_n is a random sequence. A particular test is obtained from a specifically constructed randomly weighted empirical process (RWEP) B_n. If the given sequence is from an IID sequence the process B_n behaves like a Brownian process, but quite differently for certain dependent sequences. Some optimality results are obtained in the case of contiguous ARMA alternatives. Some simulation results are also presented.

The process is also considered for residuals from regression.

ABOUT THE SPEAKER

Dr. Reg Kulperger received his Bachelor's and Master's degree from the University of Waterloo and PhD from Carleton University under Donald Dawson. He spent time as a post doctoral fellow at Berkeley, was a faculty member at McMaster for two years and then went to the University of Western Ontario where he is now Associate Professor of Statistics. His research interests are in stochastic modeling and inference for stochastic processes. A recent paper in this area deals with image correlation spectroscopy and appeared in the Canadian Journal of Statistics in 1997.

REFERENCES

The following articles have been provided by Dr. Kulperger to be used as background for his talk. They are on reserve at Thode Library (STATS 770: Statistics Seminar).

[1] Lockhart R. & Kulperger, R.J. (1998). "Tests of Independence in Time Series," JOURNAL OF TIME SERIES ANALYSIS, to appear.

[2] Kulperger R.J. (1997). "A regression Residual Process," submitted for publication.