This interview took place in Montréal, Canada.
Interviewer: Russell Barton, Ph.D.
Interview place: Montréal, Canada
Interview date: 2022-04-19
Pierre L’Ecuyer has been a professor in the Departement d’Informatique et de Recherche Opérationnelle at the Universite de Montreal since 1990. Prior to that, he was a professor in the Departement d’Informatique at Université de Laval from 1983 to 1990. He held a prestigious Canada Research Chair in Stochastic Simulation and Optimization for 2004–2017 and was an INRIA International Chair (at Inria-Rennes, France) during the period 2013–2018.
He completed his BS degree in 1972, his MS in operations research in 1980, and his Ph.D. in computer science in 1983, all at the Universite de Montreal. His Ph.D. thesis and earliest work was on Markov decision processes and approximate stochastic dynamic programming in infinite state spaces, where he contributed new theory and developed applications in economics, maintenance and reliability, production, computer systems, and computational finance.
Professor L'Ecuyer is recognized as a world authority on the design, theoretical analysis, implementation and statistical testing of random number generators (RNGs) for simulation. He has developed statistically reliable and efficient RNGs, supported by mathematical analyses of the uniformity and independence of their output values. He is also known for his work on algorithms that can provide multiple streams and substreams of random numbers in unlimited numbers. His mathematical analysis revealed significant defects in several RNGs available in software. Professor L’Ecuyer's TestU01 library is now viewed as the gold standard for statistical testing of random number generators for both simulation and cryptography.
In addition, he has made several fundamental contributions to the area of randomized quasi-Monte Carlo (RQMC) methods. These algorithms replace vectors of independent random numbers by more evenly distributed points. He has studied construction methods, proposed and compared selection criteria for the parameters, built software that constructs point sets based on these criteria, and showed how these methods can improve simulation efficiency. He has also studied applications of these ideas in finance, risk analysis, and stochastic networks. He is also known for Array-RQMC, an effective RQMC method for the simulation of Markov chains over a large number of steps, and he has shown that this method can bring large variance reductions for applications in finance and biology, for example.
L’Ecuyer is also known for his important contributions to the development and analysis of efficient methods for estimation of derivatives (or gradients) and to the development of simulation-based optimization methods. Another significant area of contribution for which he is known is variance reduction, in which an estimator is replaced by a more cleverly designed one having the same expectation but smaller variance. Some of these methods use (and sometimes combine) ideas based on conditional Monte Carlo, importance sampling, and splitting to estimate rare event probabilities.
He is a past Editor-in-Chief for ACM Transactions on Modeling and Computer Simulation, and has served as Area or Associate Editor for ACM Transactions on Mathematical Software, Management Science, Statistics and Computing, and International Transactions in Operational Research. His influence on the field has also been felt through his supervision and guidance of 23 postdoctoral fellows and 51 Ph.D. and master students.
Professor L’Ecuyer's work has been recognized with multiple prestigious awards, including the INFORMS Simulation Society Lifetime Professional Achievement Award in 2020, the Computer Science Canada Lifetime Achievement Award in 2019, the ACM SIGSIM Distinguished Contributions Award in 2016, the Canadian Operational Research Society Award of Merit in 2014, and the INFORMS Simulation Society Distinguished Service Award in 2011. His research has been recognized with the INFORMS Simulation Society Outstanding Research Publication Award three times, in 1999, 2009, and 2018, a Killam Research Fellowship in 2001-03, the Urgel-Archambault Prize from ACFAS in 2002, and a Steacie Fellowship from NSERC-Canada in 1995-97. He was selected as an INFORMS Fellow in 2006.
He has been a visiting scholar at Stanford University (USA), INRIA–Rocquencourt (France), Ecole des Mines (France), Waseda University (Tokyo), University of Salzburg (Austria), North Carolina State University (USA), Universite de Savoie in Chambery (France), University of New South Wales (Australia), several times at INRIA–Rennes (France), and Google Research (Mountain View, USA). He is currently a member of the CIRRELT and GERAD research centers in Montreal.