Sufficient Dimension Reduction in Binary Classification
Reducing dimensionality of data is essential for binary classification with high-dimensional covariates. In the context of sufficient dimension reduction (SDR), most, if not all, existing SDR methods suffer in binary classification. In this talk, we target directly at the SDR for binary classification and propose a new method based on support vector machines. The new method is supported by both numerical evidence and theoretical justification.

Bio: Yichao Wu graduated from The University of North Carolina at Chapel Hill in 2006. He is an associate professor at North Carolina State University.