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 methodological innovation and presentation. This year’s winner is the article entitled “Post model-fitting exploration via a ‘Next Door’ analysis". (Vol. 48, No. 3, 2020, Pages 447–470) by Leying Guan and Robert Tibshirani.
The authors propose an interesting and elegant approach to find and assess models “close” to a base model, a procedure that is called the “Next-Door analysis”. The idea is as follows. First the usual lasso is fit to the data. Then for each predictor in the support set, that predictor is removed predictor and the lasso is refitted to all of the remaining predictors (not just the support set). This gives a nearby model (proximal model) corresponding to the deletion of each of the member in the support set. Finally, each of these nearby models is examined and evaluated.
Leying Guan received her B.S. in physics from Tsinghua University in 2014, and a Ph.D. from the Statistics department at Stanford in 2019. She is currently an Assistant Professor at Yale Biostatistics. Her research primarily focuses on developing statistical and machine learning methods driven by scientific applications. Her recent interests include large-scale hypothesis testing, machine learning approaches under distributional shift, and statistical modeling of immunological data.
Robert Tibshirani received his B. Math. in statistics and computer science from the University of Waterloo in 1979 and a Master's degree in Statistics from University of Toronto in 1980. Tibshirani joined the doctoral program at Stanford University in 1981 and received his Ph.D. in 1984. He was a Professor at the University of Toronto from 1985-1998, and is currently a Professor in the Departments of Biomedical Data Sciences and Statistics at Stanford University. His areas of research interest include the development statistical tools for the analysis of high-dimensional and complex datasets.
The article entitled “Post model-fitting exploration via a ‘Next Door’ analysis" by Leying Guan and Robert Tibshirani is recognized for an excellent presentation of impressive methodological development and application in the analysis of complex datasets.
Andrei Volodin was primarily responsible for producing this material.