![Johanna Nešlehová Johanna Nešlehová](/sites/default/files/styles/award_winner/public/people/johanna_n.png?itok=2YShb7op)
The CRM-SSC Prize in Statistics is awarded annually by the Centre de recherches mathématiques (CRM) and the Statistical Society of Canada (SSC) in recognition of outstanding research carried out primarily in Canada by a statistician during the first fifteen years after completing a PhD. The 2019 recipient of this prize is professor Johanna Nešlehová of McGill University.
Born in Prague, Johanna is the daughter of Czech painter Pavel Nešleha and art historian Mahulena Nešlehová. She studied mathematics and statistics in Czechia (Univerzita Karlova, 1999) and Germany (Universität Hamburg, 2000; Carl von Ossietzky Universität, PhD, 2004). Her interests in multivariate analysis, nonparametric statistics and applications were stimulated by Marie Hušková, Georg Neuhaus and Dietmar Pfeifer. At ETH Zürich, where she was postdoc and later Heinz Hopf Lecturer, her expertise expanded to extreme-value theory and quantitative risk management under the guidance of Paul Embrechts. She joined McGill in 2009, was promoted to Associate Professor in 2012, and is currently Chair of the Undergraduate Programs in Mathematics and Statistics at that institution.
Since 2004, Johanna has published over 40 papers in high-caliber international journals such as Bernoulli, Biometrika, The Annals of Statistics (AoS), and the Journal of Multivariate Analysis (JMVA). In addition to making deep and lasting contributions to statistical theory and risk management, she coauthored with Erhard Cramer an undergraduate text in mathematics now in its 7th edition at Springer.
Johanna’s early papers with Paul Embrechts and others, studied the impact of extreme events on risks, developed extreme-value tools for the analysis of loss data, and critically assessed the use of infinite-mean models in operational risk. These seminal works, published in the Journal of Operational Risk and the Journal of Banking and Finance, are frequently cited.
In parallel, Johanna undertook a thorough reexamination of the prevalent dependence structures used for multivariate data modeling with copulas, most notably the Archimedean class, which extends Cox's proportional hazards model. Her groundbreaking 2009 AoS paper with Alex McNeil provided key new insights into this class of models that faciliate their use and inspired much work, including some by Johanna and her collaborators. For example in a 2011 discussion paper in TEST, she generalized a well-known rank-based estimation technique for Archimedean models. In a 2019 AoS paper, she also uses a sophisticated rank-based approach to develop semiparametric estimators of Archimax copulas which describe various forms of dependence in pre-extreme regimes.
However, Johanna’s greatest contributions are probably those concerned with the extension of rank-based copula inference techniques to the analysis of mixed data. She started investigating this issue in her PhD thesis, which led to her 2007 solo paper in JMVA. The same year, she coauthored a paper on this topic in the ASTIN Bulletin which was identified as one of the three “must-read” of copula modeling in the Journal of Risk and Insurance. She is actively pursuing this line of research, mostly with Christian Genest and Bruno Rémillard. Noteworthy are her 2014 Bernoulli and 2017 JMVA papers, which make heavy use of the theory of empirical processes to resolve the thorny issue of ties in validating rank-based inference procedures for mixed data. A 2019 Biometrika paper of hers exploits these results to develop powerful tests of independence for sparse contingency tables whose dimension increases with sample size. She has also devised techniques for detecting patterns in large-scale correlation matrices.
Beyond her great productivity in research, which includes applications in environmental and health sciences, Johanna has an outstanding record of graduate training. She is also an exceptionally inspiring lecturer who receives frequent invitations to speak at international meetings. She has served the community in many ways, foremost as an Associate Editor for journals such as JMVA and The Canadian Journal of Statistics, but also on committees for the SSC and Bernoulli Society, and as an organizer for two thematic semesters at the CRM. She was elected a member of the International Statistical Institute in 2011 and held a John von Neumann Guest Professorship at Technische Universität München in 2016. Outside work, she has a keen interest in art and history; she also enjoys skiing and family activities with her husband Christian and their son Richard.
Johanna will present an overview of her work in a Special Session at the 47 th Annual Meeting of the Statistical Society of Canada to be held in Calgary, Alberta, May 26 to 29, 2019.
“To Johanna Nešlehová for fundamental contributions to multivariate statistics, and in particular stochastic dependence modeling and extreme-value theory, and for her efforts to promote the sound application of statistics in risk management.”