On Thursday October 16th at 15h30, Prof. Arthur Zimek, Southern Denmark University will give a seminar at the amphitheater E1, building E (St. Martin). The presentation is expected to last about 45min. Professor Zimek is visiting ETIS for the period of 14-17 October; do not hesitate to contact him if you wish to meet with him.
Abstract: Fairness and bias issues in classification are particularly prevalent when the numbers of examples for different classes are out of proportion. In machine learning this is known as the problem of imbalanced classification. While it is well known that recall rather than precision is the performance measure to optimize in imbalanced classification problems, most existing methods that adjust for class imbalance do not particularly address the optimization of recall. In this talk, we discuss an elegant and straightforward variation of the k-nearest-neighbor classifier to balance imbalanced classification problems internally in a probabilistic interpretation and show how this relates to the optimization of the recall.
Arthur Zimek, Southern Denmark University, Dr. rer. nat. habil.
Head of Section, Professor, Data Science
Department of Mathematics and Computer Science
University of Southern Denmark (SDU)
Arthur Zimek is Full Professor and Head of the Data Science and Statistics section in the Department for Mathematics and Computer Science (IMADA) at University of Southern Denmark (SDU), in Odense, Denmark. Previous positions were are LMU Munich, Germany, TU Wien, Austria, and University of Alberta, Edmonton, Canada. Several awards include the ”SIGKDD Doctoral Dissertation Award (runner-up)” in 2009, the ”Best Demonstration Paper Award” at SSTD 2011, the ”Best Research Paper Award” at SDM 2024, and a listing in the ACM Computing Reviews ”21st Annual Best of Computing” (2016). His research interests include ensemble techniques for unsupervised learning, clustering, outlier detection, high dimensional data, and explainable AI, developing data mining methods as well as evaluation methodology. He serves as associate editor for the Springer Data Mining and Knowledge Discovery journal and the Springer Machine Learning journal.
