用酉变换近似矩阵和张量对角化：Jacobi型算法的收敛性
Approximate matrix and tensor diagonalization by unitary transformations: convergence of Jacobi-type algorithms
摘要：
针对酉群上的一类极大化问题，提出了一种基于梯度的Jacobi算法，重点研究了利用酉变换对复矩阵和张量进行近似对角化.给出了算法的弱收敛结果，并证明了算法的局部线性收敛，收敛结果也适用于实值张量的情形。
英文摘要：
We propose a gradient-based Jacobi algorithm for a class of maximization problems on the unitary group, with a focus on approximate diagonalization of complex matrices and tensors by unitary transformations. We provide weak convergence results, and prove local linear convergence of this algorithm.The convergence results also apply to the case of real-valued tensors.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
