A0752
Title: Dimension reduction through imbalanced learning
Authors: Qin Wang - The University of Alabama (United States) [presenting]
Abstract: Sufficient dimension reduction (SDR) is a useful tool in high-dimensional data analysis. It aims to find informative embedding subspaces without losing regression information. Inverse regression-based methods, including SIR and SAVE, have been proposed and well-studied in the literature. A resampling-based approach through imbalanced learning is proposed to further enhance estimation accuracy and consistency and ease the challenge of selecting the number of slices using traditional SDR approaches. Numerical studies and a real data application will be presented to illustrate the efficacy of the proposed method.
