报告题目：Structural Sparsity from Differential Inclusion to Deep Learning
报告摘要：We consider recovering the signal with structural sparsity, i.e., the signal after a linear transformation is sparse. In this talk, we introduce a differential inclusion approach with variable splitting mechanism, which generates a regularization solution path. Equipped with the variable splitting, our method can alleviate the multicollinearity problem. Under this condition, we can prove the model selection consistency for this solution path. We further propose a data adaptive approach to determine a proper early stopping mechanism, towards controlling False-Discovery -Rate of selected features. Finally, we apply our method to explore the sparse structure of deep networks. Guided by differential inclusion, our method can forwardly learn the important weights/filters from sparse to dense, which can largely reduce the computational cost. Theoretically, we can prove the global convergence of this learning process.
Dr. Xinwei Sun is now a researcher in Microsoft Research Asia (MSRA). He obtained PhD degree in statistics from Peking University in 2018. His research interests focus on the statistical models and their applications on medical imaging analysis. He has published over 20 papers in top mathematical journals such as ACHA, machine learning conferences such as NeurIPS, ICML, medical imaging-related conferences such as MICCAI, CVPR, ICCV, ECCV, etc. Many of his works on medical imaging analysis have been transformed into medical products.