The Gumbel Lecture, which was held for the first time at the Statistical Week 2006, bears the name of the German statistician Emil Julius Gumbel.
As a population statistician and founder of extreme value statistics, Gumbel is one of the most important representatives of his field. Born in Munich in 1891, he studied mathematics and economics there. After graduating in actuarial mathematics, he worked as an assistant to Georg von Mayr, founder of the “Allgemeines Statistisches Archiv” and first Chairman of the German Statistical Society. Gumbel received his doctorate under Mayr in 1914 with his paper “Die Berechnung des Bevölkerungsstandes durch Interpolation.” In the course of World War I, Gumbel became an active pacifist. While working in Berlin at Telefunken, he came into contact with, among others, Albert Einstein and Ladislaus von Bortkiewicz. Influenced by the latter, he turned to probability-theoretical modeling of statistics, especially the creation of mortality tables. In 1923 he was habilitated in Heidelberg for “Mathematical Statistics”. Because of his political publications, especially on the femicide, the Nazi students at Heidelberg University demanded his dismissal, which finally took place in the summer of 1932. He was among the first academics expelled by the Nazis. The “expatriation” took place in August 1933 (1st expatriation list). With the rise of the Third Reich, he had to leave Germany – due to his Jewish origin and his pacifist activities. Gumbel was a visiting professor at the Institut Pointcaré in Paris in the winter of 1932/33. Thanks to the support of his French colleagues, he got a position as maitre des recherches at the Institut de Science Financière te d’Assurances of the University of Lyon in 1934. In 1940 he had to flee again. Thanks to an affidavit from the New School for Social Research in New York, he managed to escape to the United States. He worked at the New School from 1940 to 1945, after which he taught at various colleges in New York and received research assignments. From 1953 he was an adjunct professor in the Department of Engineering at Columbia University New York. Emil Julius Gumbel died in New York in 1966.
In the 1930s, Gumbel developed the foundations of extreme value statistics. Starting from the actuarial problem of life tables and life distributions, he studied the maximum of independent random variables and derived their limiting distributions. The most important writings of this period are “Das Zufallsgesetz des Sterbens” (The Random Law of Dying) of 1932 and the monograph “La durée extrême de la vie humaine” of 1937. He applied his theoretical findings not only in actuarial science, but also in climate research – specifically in the prediction of floods and other climatic extremes. He published his major work in 1958 in the monograph “Statistics of Extremes”.
A detailed account of Gumbel’s life and work can be found in Rendtel, U.; Wasmuht, U.; Wilrich, P.-Th. (2021): Emil Julius Gumbel – Innovative Statistician and Committed Pacifist. AStA Wirtschafts- und Sozialstatistisches Archiv, Online, https://doi.org/10.1007/s11943-021-00293-0.
Overview
Year | Title |
---|---|
2021 | Prof. Dr. Sonja Greven, Elastic analysis of irregularly or sparsely sampled curves |
2019 | Prof. Dr. Axel Bücher, On the block maxima method in extreme value statistics |
2018 | Prof. Dr. Tatyana Krivobokova, Adaptive estimation and testing in nonparametric regression with dependent errors |
2017 | Prof. Dr. Carsten Jentsch, Statistical inference on party positions from texts: statistical modeling, bootstrap and adjusting for time effects |
2016 | Prof. Dr. Hajo Holzmann, Nonparametric Identification and Estimation in a Triangular Random Coefficient Regression Model |
2015 | Prof. Dr. Christoph Rothe, Partial identication in regression discontinuity designs with manipulated running variables |
2014 | Prof. Dr. Alexander Aue, The Marcenko-Pastur Law for Linear Time Series |
2013 | Prof. Dr. Mark Podolskij, Inference and limit theory for ambit fields |
2012 | Prof. Dr. Davy Paindaveine, Nonparametrically consistent depth-based classifiers |
2011 | Prof. Dr. Natalie Neumeyer, Specification testing in nonparametric quantile regression |
2010 | Prof. Dr. Patrik Guggenberger, On the Asymptotic Size of Testing Procedures |
2009 | Prof. Dr. Martin Wagner, Cointegration Analysis in State Space Systems |
2008 | Prof. Dr. Michael Wolf, Recent Advances in Multiple Testing |
2007 | Prof. Dr. Evgueni Spodarev, Räumliche Risikoanalyse in der Schadensversicherung |
2006 | Prof. Dr. Cristiano Varin, On composite marginal likelihood inference for hierarchical models |