Next AI Colloquium with Prof. Dr. Arne Bathke!

When: 19th September 2024, 09:00-10:00am

Where: Joseph-von-Fraunhofer-Straße 25, room 303 and via Zoom 

Who: Prof. Dr. Arne Bathke from the University of Salzburg

Title: Inference methods for multivariate data

The AI Colloquium is co-organized by the Lamarr Institute for Machine Learning and Artificial Intelligence, the Research Center Trustworthy Data Science and Security (RC Trust), and the Center for Data Science & Simulation at TU Dortmund University (DoDaS).

Short Bio:

Born in Hamburg, grew up in Northern Germany and Norway. Studied mathematics in Göttingen, spent study time abroad in Italy and the USA, and obtained his doctoral degree in 2000. Afterwards, has worked for 11 years in the United States as statistics professor at the University of Kentucky, where he was also inaugural director of the Applied Statistics Laboratory, a comprehensive statistics consulting facility on campus. Since 2012 statistics professor at the University of Salzburg where he has also been elected to different administrative positions (Dean since 2015). Among other initiatives, first Data Science MSc program at an Austrian university (2016) and MINT:labs (2018) where school children can experience MINT (=STEM) topics in hands-on workshops at the university.

Arne Bathke’s research deals with developing and evaluating new statistical methods, and he has given numerous invited talks, lectures and workshops on five continents. On the other hand, he is also interested in the application and improvement of statistical procedures in interdisciplinary cooperations with colleagues from other fields, from medicine, biology, and ecology to education and economics, as well as in the Austrian Future Operations Platform. About half of his more than 100 publications in international journals result from such cooperation projects.

Among other awards, Arne Bathke is a Fellow of the American Statistical Association and an Elected Member of the European Academy of the Sciences and the Arts. He has been named “Henry Clay Ambassador” by the Mayor of Lexington (Kentucky) for his civil engagement, and he has received two university-wide awards for excellence in teaching and advising. Furthermore, he has been President of the International Biometric Society - Region Österreich-Schweiz (IBS-ROeS), is on the board of the Austrian Statistical Association (ÖSG), Editor-in-Chief of Biometrical Journal, the flagship journal of the International Biometric Society Regions Germany, Austria-Switzerland, and Italy, as well as on the editorial board of three other international statistics journals (International Journal of Biostatistics, Journal of the American Statistical Association, The American Statistician). Organizer of an annual Biostatistics Summer School, and has organized and co-organized several meetings, workshops, and conferences.

Abstract:

When there are multiples response variables (endpoints) and different predictors, one typically wants to find out which predictors are relevant, and for which endpoints. We present two rather general approaches towards valid inference for multivariate data, one accommodating binary, ordinal, and metric endpoints, and the other allowing for a factorial design structure. The first approach is fully nonparametric, resulting in rank-based statistics and an F-approximation of the sampling distribution, while the second approach employs asymptotically valid resampling techniques (bootstrap). We also try to address the question of how well the proposed methods actually accomplish their goals, and how to use the respective toolboxes that have been developed.



We look forward to seeing you.