Fast truncation of mode ranks for bilinear tensor operations

  title={Fast truncation of mode ranks for bilinear tensor operations},
  author={Dmitry V. Savostyanov and Eugene E. Tyrtyshnikov and Nickolai Zamarashkin},
  journal={Numer. Linear Algebra Appl.},
SUMMARY We propose a fast algorithm for mode rank truncation of the result of a bilinear operation on 3-tensors given in the Tucker or canonical form. If the arguments and the result have mode sizes n and mode ranks r, the computation costs (nr3 + r4). The algorithm is based on the cross approximation of Gram matrices, and the accuracy of the resulted Tucker approximation is limited by square root of the machine precision. We apply the proposed algorithms for the evaluation of the Hadamard… Expand
Fast Multidimensional Convolution in Low-Rank Tensor Formats via Cross Approximation
A new cross-conv algorithm for approximate computation of convolution in different low-rank tensor formats (tensor train, Tucker, hierarchical Tucker), based on applying cross approximation in the “frequency domain,” where convolution becomes a simple elementwise product. Expand
Fast multidimensional convolution in low-rank formats via cross approximation
We propose a new cross-conv algorithm for approximate computation of convolution in different low-rank tensor formats (tensor train, Tucker, Hierarchical Tucker). It has better complexity withExpand
CANDECOMP/PARAFAC Decomposition of High-Order Tensors Through Tensor Reshaping
By exploiting the uniqueness of CPD and the relation of a tensor in Kruskal form and its unfolded tensor, a fast approach to deal with this problem is proposed and strategies to unfold tensors are suggested and practically verified. Expand
Numerical tensor calculus*
The algebraic and analytical aspects of tensor spaces are introduced and the main part concerns the numerical representation of tensors and the numerical performance of Tensor operations. Expand
Superfast Fourier Transform Using QTT Approximation
We propose Fourier transform algorithms using QTT format for data-sparse approximate representation of one- and multi-dimensional vectors (m-tensors). Although the Fourier matrix itself does not haveExpand
A literature survey of low‐rank tensor approximation techniques
This survey attempts to give a literature overview of current developments in low-rank tensor approximation, with an emphasis on function-related tensors. Expand
DMRG Approach to Fast Linear Algebra in the TT-Format
  • I. Oseledets
  • Mathematics, Computer Science
  • Comput. Methods Appl. Math.
  • 2011
The concept of the DMRG minimization scheme is extended to several important operations in the TT-format, like the matrix-by-vector product and the conversion from the canonical format to the TT -format. Expand
Tensor Clustering: A Review
This survey concisely presents the key tensor clustering techniques as well as their applications and deep learning frameworks which natively support tensors such as TensorFlow, Breeze, Spark MLlib, and Tensor Toolbox are presented. Expand
Structured matrices and tensors
The Third International Conference on Structured Matrices and Tensors was held from January 19 to January 22, 2010 at Hong Kong Baptist University, The Hong Kong Polytechnic University, and Chinese University of Hong Kong, with the aim in mind that emerging applications will stimulate new theoretical research and that better theoretical tools in turn can be exported back to various fields of application. Expand


Wedderburn Rank Reduction and Krylov Subspace Method for Tensor Approximation. Part 1: Tucker Case
A matrix approximation algorithm that computes the Krylov subspaces and can be generalized to 3-tensors is proposed and the numerical experiments show that on quantum chemistry data the proposed tensor methods outperform the minimal Krylov recursion. Expand
Black Box Low Tensor-Rank Approximation Using Fiber-Crosses
AbstractIn this article we introduce a black box type algorithm for the approximation of tensors A in high dimension d. The algorithm adaptively determines the positions of entries of the tensor thatExpand
Multigrid Accelerated Tensor Approximation of Function Related Multidimensional Arrays
A novel tensor approximation method for discretized multidimensional functions and operators in $\mathbb{R}^d$, based on the idea of multigrid acceleration, which indicates linear complexity with respect to the size of data representing the input tensor. Expand
On the Best Rank-1 and Rank-(R1 , R2, ... , RN) Approximation of Higher-Order Tensors
A multilinear generalization of the best rank-R approximation problem for matrices, namely, the approximation of a given higher-order tensor, in an optimal least-squares sense, by a tensor that has prespecified column rank value, rowRank value, etc. Expand
Linear algebra for tensor problems
This paper presents fast recompression procedures with sublinear complexity with respect to the size of data and proposes methods for basic linear algebra operations with all matrix operands in the Tucker format, mostly through calls to highly optimized level-3 BLAS/LAPACK routines. Expand
Cross approximation in tensor electron density computations
The Tucker format has advantages in the storage and computation time compared with the ubiquitous canonical format when dealing with Hartree–Fock/Kohn–Sham equations. Expand
Tensor Rank and the Ill-Posedness of the Best Low-Rank Approximation Problem
It is argued that the naive approach to this problem is doomed to failure because, unlike matrices, tensors of order 3 or higher can fail to have best rank-r approximations, and a natural way of overcoming the ill-posedness of the low-rank approximation problem is proposed by using weak solutions when true solutions do not exist. Expand
Tensor decomposition in electronic structure calculations on 3D Cartesian grids
It is shown that the orthogonal Tucker-type tensor approximation of electron density and Hartree potential of simple molecules leads to low tensor rank representations, which enables an efficient tensor-product convolution scheme for the computation of the Hartrees potential using a collocation-type approximation via piecewise constant basis functions on a uniform nxnxn grid. Expand
Computation of the Canonical Decomposition by Means of a Simultaneous Generalized Schur Decomposition
The canonical decomposition of higher-order tensors is a key tool in multilinear algebra and, under certain conditions, the problem can be rephrased as the simultaneous diagonalization, by equivalence or congruence, of a set of matrices. Expand
Tensor Decompositions, State of the Art and Applications
  • P. Comon
  • Mathematics, Computer Science
  • 2002
A partial survey of the tools borrowed from tensor algebra, which have been utilized recently in Statistics and Signal Processing, shows why the decompositions well known in linear algebra can hardly be extended to tensors. Expand