Non-convex Factorization and Manifold Formulations

讲座简介

In this talk, we consider the geometric landscape connection of the widely studied manifold and factorization formulations in low-rank positive semidefinite (PSD) and general matrix optimization. We establish an equivalence on the set of first-order stationary points (FOSPs) and second-order stationary points (SOSPs) between the manifold and the factorization formulations. We further give a sandwich inequality on the spectrum of Riemannian and Euclidean Hessians at FOSPs, which can be used to transfer more geometric properties from one formulation to another. We also discuss applications of our findings to some machine learning problems.

时间

021-12-01

下午 15:00 ~ 16:30

主讲人

郦旭东，复旦大学大数据学院、上海数学中心双聘青年研究员

地点

会议号码:  613 690 699

参会密码:  123456

参会链接：https://meeting.tencent.com/dm/GwYrBupjb4Xv