On the Nuclear Norm and the Singular Value Decomposition of Tensors

被引:27
作者
Derksen, Harm [1 ]
机构
[1] Univ Michigan, Dept Math, 530 Church St, Ann Arbor, MI 48109 USA
基金
美国国家科学基金会;
关键词
Tensor decomposition; Nuclear norm; Singular value decomposition; PARAFAC; CANDECOMP; MATRIX MULTIPLICATION; RANK; UNIQUENESS; COMPLEXITY;
D O I
10.1007/s10208-015-9264-x
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
Finding the rank of a tensor is a problem that has many applications. Unfortunately, it is often very difficult to determine the rank of a given tensor. Inspired by the heuristics of convex relaxation, we consider the nuclear norm instead of the rank of a tensor. We determine the nuclear norm of various tensors of interest. Along the way, we also do a systematic study various measures of orthogonality in tensor product spaces and we give a new generalization of the singular value decomposition to higher-order tensors.
引用
收藏
页码:779 / 811
页数:33
相关论文
共 36 条
[1]  
Aliprantis CD., 2006, INFINITE DIMENSIONAL
[2]  
[Anonymous], 1955, Memoirs Amer. Math. Soc
[3]   O(N2.7799) COMPLEXITY FOR N BY N APPROXIMATE MATRIX MULTIPLICATION [J].
BINI, D ;
CAPOVANI, M ;
ROMANI, F ;
LOTTI, G .
INFORMATION PROCESSING LETTERS, 1979, 8 (05) :234-235
[4]   Beyond the Alder-Strassen bound [J].
Bläser, M .
THEORETICAL COMPUTER SCIENCE, 2005, 331 (01) :3-21
[5]  
Burgisser P., 1997, ALGEBRAIC COMPLEXITY, V315
[6]   The Power of Convex Relaxation: Near-Optimal Matrix Completion [J].
Candes, Emmanuel J. ;
Tao, Terence .
IEEE TRANSACTIONS ON INFORMATION THEORY, 2010, 56 (05) :2053-2080
[7]   Exact Matrix Completion via Convex Optimization [J].
Candes, Emmanuel J. ;
Recht, Benjamin .
FOUNDATIONS OF COMPUTATIONAL MATHEMATICS, 2009, 9 (06) :717-772
[8]  
Carlen E, 2006, METHODS APPL ANAL, V13, P1
[9]  
Carroll JD, 1970, PSYCHOMETRIKA, V35, P218
[10]   MATRIX MULTIPLICATION VIA ARITHMETIC PROGRESSIONS [J].
COPPERSMITH, D ;
WINOGRAD, S .
JOURNAL OF SYMBOLIC COMPUTATION, 1990, 9 (03) :251-280