Trevor Hastie, professor at Stanford University, is one of the outstanding statisticians of our time. His research contributions lie in applied statistics and in the field of statistical learning, which make him one of the pioneers of Big Data analysis. His research focuses on modelling and prediction problems in biology and genetic research, medicine and industry. He has published more than 200 articles and written six highly regarded monographs. These include the standard textbook ‘Elements of Statistical Learning’ published together with Rob Tibshirani and Jerome Friedman. As in many other universities, his works constitute an indispensable part of the literature canon in the Leuphana Data Science Programme.
Two fundamental fields of his work deserve special mention. First: Real-world phenomena are often described using non-linear models. So-called Generalised Additive Models, to whose development he made a decisive contribution, offer an approach combining a high degree of flexibility in modelling with the many advantages of linear models. In this context, his early work on parameter-free models is highly noteworthy. Second: In times of big data, the number of possible predictors often exceeds that of observations. Trevor Hastie and his colleagues have developed a number of methods to select relevant predictors in this situation.
Trevor Hastie was born in South Africa in 1953. He studied at Rhodes University and Cape Town University. He received his doctorate from Stanford University in 1984. He joined AT&T Bell Laboratories in 1986 and in 1994 returned to Stanford University as Professor of Statistics and Biostatistics. In 2013, he was appointed John A. Overdeck Professor of Mathematics.