题目： Closed-Form Solutions in Low-Rank Subspace Recovery Models and Their Implications
报告人： 林宙辰 教授 北京大学
摘要： In recent years, in parallel to sparse models, low-rank models also develop significantly, in particular in subspace recovery problems. Besides robustness, low-rank subspace recovery models may have closed form solutions. This property further makes low-rank models distinct from sparse models. The results are both surprising and elegant. Moreover, they also open doors to better solutions and faster numerical algorithms. In this talk, I will present my recent work in this line.