Quadratic tensor eigenvalue complementarity problems

报告摘要：In this talk, we discuss how to solve the quadratic tensor eigenvalue complementarity problem (QTEiCP). By a randomization process, the quadratic complementarity (QC) eigenvalues are classified into two cases. For each case, the QTEiCP is formulated as an equivalent generalized moment problem. The QC eigenvectors can be computed in order. Each of them can be solved by a sequence of semidefinite relaxations. We prove that such a sequence converges in finitely many steps for generic tensors.

报告人简介：范金燕，上海交通大学数学科学学院教授。主要从事非线性最优化的理论和方法研究，在非线性方程组、完全正优化、张量计算等方面取得了一系列成果。现为J. Ind. Manag. Optim.、J. Oper. Res. Soc. China、《计算数学》等期刊编委。曾获中国青年女科学家奖、中国青年科技奖，入选国家“万人计划”科技创新领军人才。