Name: Dawei Zhou
Date/Time: Thursday February 11 @ 11:00 am
Title: Learning More from Less: Complex Rare Category Analysis
Abstract: In contrast to the sheer volume of data being collected in the age of big data, it is often the rare categories (i.e., the minority classes) that are of great importance in many high impact domains, ranging from financial fraud detection to rare disease diagnosis, from national security to scientific discovery. However, modern AI systems typically require the availability of rich annotated data and often achieve suboptimal performance in rare category analysis due to the label scarcity nature of rare examples. In this talk, I will discuss my recent work on complex rare category analysis, focusing on two major challenges, i.e., how to characterize the under-represented patterns with a compact representation, and how to provide the proper lens for the end-users to draw insights from the data and rare category analysis models. In particular, I will hinge on key application domains, discuss our proposed techniques and theoretical results for characterizing and comprehending rare examples, and showcase a unified visual analytic system for rare category analysis in the dynamic environment. Finally, I will conclude this talk and share thoughts about my future plans.
Bio: Dawei Zhou is a final year Ph.D. candidate at the Computer Science Department, University of Illinois at Urbana-Champaign, under the supervision of Dr. Jingrui He. His research interests lie in rare category analysis, graph mining, curriculum learning, and algorithmic fairness, with applications in financial fraud detection, financial forecasting, social media analysis, and healthcare. He has authored more than 20 publications in premier academic venues across AI, data mining, and information retrieval (e.g., AAAI, IJCAI, KDD, ICDM, SDM, TKDD, DMKD, WWW, CIKM) and has received student travel awards at KDD, WWW, AAAI, IJCAI, ICDM, etc. His work on complex rare category analysis has been selected by Computing Research Association (CRA) to showcase at the 24th CNSF Capitol Hill Science Exhibition. He has broad collaborations within industry and academia, such as IBM T.J. Watson Research Lab, HRL Laboratories, Alibaba DAMO Academy, Early Warning Inc., ASU, University of Rochester, Rutgers University, etc.