A Levenberg-Marquardt Method for Tensor Approximation

This paper presents a tensor approximation algorithm, based on the Levenberg–Marquardt method for nonlinear least square problem, to decompose large-scale tensors into sum of products vector groups given scale, or obtain low-rank without losing too much accuracy. An Armijo-like rule inexact line search is also introduced this algorithm. The result decomposition adjustable, which implies that can be specified according users’ requirements. convergence proved, and numerical experiments show it has some advantages over classical method. algorithm applicable both symmetric asymmetric tensors, expected play role in field data compression storage operations.