Jerry Sacks and the Birth of Design and Analysis of Computer Experiments

William Welch Speaker
University of British Columbia
Sunday, Aug 6: 2:00 PM - 3:50 PM
Invited Paper Session 
Metro Toronto Convention Centre 
We will review in three parts Jerry's central role in the creation of a new, and eventually enormous, field of research and application: the Design and Analysis of Computer Experiments (DACE). (1) Jerry, in collaboration with Don Ylvisaker, had a remarkable series of papers on statistical design to make regression robust to nonparametric, correlated errors. With adaptation it became the underlying Gaussian process paradigm for DACE. (2) Jerry played a pivotal role in recognizing and proposing a research program for DACE, along with Toby Mitchell, Henry Wynn, and Don Ylvisaker. We will briefly discuss how that came about. (3) Starting in the late 1980s, Jerry had the central role in a series of seminal papers on DACE, coauthored with the aforementioned researchers and others. Those papers established a methodology for analysis of computer experiments which is still ubiquitous today and the basis for efficient statistical design of computer experiments. Thus, the talk will recount some of the story of how Jerry developed and adapted statistical tools for the scientific revolution where physical experiments are complemented or supplanted by computer models of complex systems.