主讲人简介：王莉博士，美国伊利诺伊芝加哥大学，研究助理教授。2014年获美国加州大学圣地亚哥分校应用数学专业博士学位，毕业后在美国布朗大学和加拿大维多利亚大学做博士后一年。现在美国伊利诺伊芝加哥大学做研究助理教授。主要研究方向：最优化理论与算法，多项式规划，大数据，机器学习等。在《SIAM  Journal on Optimization》，《SIAM Journal on Matrix Analysis and  Applications》以及《IEEE Transactions on Signal  Processing》等期刊上发表论文12篇，以及会议上发表论文7篇。

内容摘要：Polynomial optimization is a special nonlinear programming in which both  the objective and constraints are polynomials. This kind of problem is always  NP-hard even if the objective is nonconvex quadratic and all constraints are  linear. The semidefinite (SDP) relaxations approach, based on sum of squares  representations, provides us with strong tools to solve polynomial optimization  problems with finitely many constraints globally. My talk will give an  introduction about this approach and some applications of this approach will be  discussed