摘要
首先定义低阶张量乘法和4阶张量的逆,考察4阶张量的可逆条件及其Tucker分解,引入多重混合线性张量(MMLT)模型,最后结合张量CP分解和Tucker分解给出MMT模型的最优参数估计。
We first defined the low order tensor multiplication and the inverse of a 4-order hypercubic tensor, and investigated the condition for a 4-order tensor to be invertible and the Tucker decomposition of a 4^th order tensor. Then we introduced a multivariate mixed tensor (MMLT) model. Finally,the parameter estimation of a multivariate linear mixed tensor model was given by the CP decomposition and Tucker decomposition of the design tensor of the model.
出处
《苏州科技大学学报(自然科学版)》
CAS
2017年第1期42-48,共7页
Journal of Suzhou University of Science and Technology(Natural Science Edition)
基金
国家自然科学基金重大项目(61190114/F0102)
香港政府研究基金项目(Poly U 501212
501913
15302114
15302115)