Lajos Horváth, Piotr Kokoszka and Matthew L. Reimherr, The Canadian Journal of Statistics Award 2010
The Canadian Journal of Statistics Award is presented each year by the Statistical Society of Canada to the author(s) of an article published in the Journal, in recognition of the outstanding quality of the paper’s methodological innovation and presentation.
This year’s winner is the article entitled “Two sample inference in functional linear models” (vol. 37, no 4, pp. 571-91), by Lajos Horváth, Piotr Kokoszka, and Matthew L. Reimherr. The paper proposes a method of comparing two linear models in which explanatory variables are functions and response variables can be either scalars or functions. In such models, the parameter vectors (or matrices) are integral operators acting on a function space. The authors show how to test whether these operators are the same in two independent samples. Their tests are based on statistics whose limiting distribution is always chi-squared, whether the response variables are scalars or functions. In addition to having good finite-sample properties, these statistics are readily computable using the R package FDA. In the paper, the approach is illustrated using egg-laying curves of Mediterranean flies and data from terrestrial magnetic observatories.
Lajos Horváth was born, raised and educated in Hungary. Before joining the University of Utah, he was a postdoctoral fellow at Carleton University from 1985 to 1987 under the supervision of Miklós Csörgő. Lajos is a very prolific researcher. The author or co-author of three books and some 250 papers, he is a specialist in asymptotic theory and its applications in the context of time series, empirical processes, and functional data analysis.
A Professor in the Department of Mathematics and Statistics at Utah State University in Logan, Piotr Kokoszka is originally from Poland. He studied applied mathematics at Wrocław Technical University and probability theory at Boston University. He was a postdoctoral fellow in Salt Lake City and a lecturer at the University of Liverpool before taking his current position in 2000. His research is concerned with stochastic modeling and statistical inference for time series, functional data, and applications to space physics.
Matthew L. Reimherr was born and raised in Salt Lake City. He received a B.S. in Mathematics and an M.S. in Statistics at the University of Utah, under the supervision of Lajos Horváth. He is currently a Ph.D. student in the Department of Statistics at the University of Chicago. His interests are in time series, functional data analysis, and statistical genetics.