Abstract
The importance of using proper classification metrics in biomarker evaluation has been severely overlooked in scientific research. This talk first covers some key concepts including classification types and classification metrics in multi-class classification from a statistician's perspective. Then I will address some common pitfalls in multi-class classification and highlight the importance of matching classification metrics to the research aims. Finally, I will present some newly developed metrics for multi-class classification when sub-classes are involved, as well as related statistical inference methods.
Department students and members are invited to meet with Dr. Tian before the presentation. Sign up for your small-group appointment here.
Lili Tian, PhD, is a professor, associate chair, and director of graduate studies in the Department of Biostatistics at the University at Buffalo, State University of New York. Dr. Tian is also an adjunct professor of oncology at Roswell Park Comprehensive Cancer Institute. Her research interests include classification metrics, biomarker evaluation/combination, and generalized inference. She has published extensively in the fields of medical diagnosis, biomarker evaluation, and exact inference, as well as in the domains of cancer research, epidemiology, and nursing research. Dr. Tian is an elected ASA fellow and currently serves as co-editor-in-chief for Statistical Methods in Medical Research.