Dynamical Approximation by Hierarchical Tucker and Tensor-Train Tensors
Open accessThe DOI matches its Crossref record; 31 of the 31 references it cites were found in a registry or an index.
Abstract
We extend results on the dynamical low-rank approximation for the treatment of time-dependent matrices and tensors (Koch and Lubich; see [SIAM J. Matrix Anal. Appl., 29 (2007), pp. 434--454], [SIAM J. Matrix Anal. Appl., 31 (2010), pp. 2360--2375]) to the recently proposed hierarchical Tucker (HT) tensor format (Hackbusch and Kühn; see [J. Fourier Anal. Appl., 15 (2009), pp. 706--722]) and the tensor train (TT) format (Oseledets; see [SIAM J. Sci. Comput., 33 (2011), pp. 2295--2317]), which are closely related to tensor decomposition methods used in quantum physics and chemistry. In this dynamical approximation approach, the time derivative of the tensor to be approximated is projected onto the time-dependent tangent space of the approximation manifold along the solution trajectory. This approach can be used to approximate the solutions to tensor differential equations in the HT or TT format and to compute updates in optimization algorithms within these reduced tensor formats. By deriving and analyzing the tangent space projector for the manifold of HT/TT tensors of fixed rank, we obtain curvature estimates, which allow us to obtain quasi-best approximation properties for the dynamical approximation, showing that the prospects and limitations of the ansatz are similar to those of the dynamical low rank approximation for matrices. Our results are exemplified by numerical experiments.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“For the purpose of this paper, the TDVP can most easily be understood in a geometric fashion, namely as an orthogonal projection of the evolution vector of the Schrödinger equation (-iH | ) onto the tangent space of the MPS manifold M MPS at the current state”
From Unifying time evolution and optimization with matrix product states (2016) · cites this paper as [19]
In the citing paper
“The key ingredient is to project the right-hand side of the Schrödinger equation, Ĥ | , onto the tangent space, so that the evolution never leaves the manifold.”
From Unifying time evolution and optimization with matrix product states (2016) · cites this paper as [19]
In the citing paper
“An alternative idea for time evolution, the Dirac-Frenkel time-dependent variational principle (TDVP), was formulated only recently for the variational class of MPS [18,19].”
From Unifying time evolution and optimization with matrix product states (2016) · cites this paper as [19]
In the citing paper
“The TDVP equations define a simultaneous update for all MPS tensors and, as a consequence thereof, its original implementation depends on inverses of matrices conditioned by the Schmidt coefficients, similar to the original TEBD implementation.”
From Unifying time evolution and optimization with matrix product states (2016) · cites this paper as [19]
Citations
Cites
Tensor Decompositions and Applications
Kolda, T. G., & Bader, B. W. (2009) · SIAM Review · doi:10.1137/07070111x · Cited by 10,301 · cited 4×
“, [17]) C((1)) = A1B ((1))(A3 ⊗A2) , C((2)) = A2B((2))(A3 ⊗A1) .”
BackgroundInfluentialThe density-matrix renormalization group in the age of matrix product states
Schollwöck, U. (2011) · Annals of Physics · doi:10.1016/j.aop.2010.09.012 · Cited by 4,818 · cited 2×
“Formats of this type have been used successfully in the quantum physics and chemistry communities in the last decade [21, 23, 27, 32].”
MethodOseledets, I. V. (2011) · SIAM Journal on Scientific Computing · doi:10.1137/090752286 · Cited by 2,636 · cited 5×
“Typical choices for such dimension trees are those obtained by a balanced splitting of the index set, as used, e.g., by Grasedyck [7], or a front-to-back splitting which (with minor adaptations; see Remark 1) leads to the TT format due to Oseledets and Tyrtyshnikov [25, 26].”
BackgroundMethodInfluentialBeck, M. (2000) · Physics Reports · doi:10.1016/s0370-1573(99)00047-2 · Cited by 2,462 · cited 1×
“(ii) Although the HT/TT formats were only recently introduced in the numerical analysis community, they have—similar to the parallels between the Tucker format and the formulation of the multi-configuration time-dependent Hartree (MCTDH) method in quantum chemistry [5, 4, 22]—a considerable significance in various fie…”
Method(1991) · Mathematics and Computers in Simulation · doi:10.1016/0378-4754(91)90038-5 · Cited by 2,437
The multi-configurational time-dependent Hartree approach
Meyer, H.-D., Manthe, U., & Cederbaum, L. S. (1990) · Chemical Physics Letters · doi:10.1016/0009-2614(90)87014-i · Cited by 2,013 · cited 1×
“(ii) Although the HT/TT formats were only recently introduced into the numerical analysis community, they have—similar to the parallels between the Tucker format and the formulation of the multiconfiguration time-dependent Hartree (MCTDH) method in quantum chemistry [5, 4, 22]—a considerable significance in various fi…”
MethodOptimization Algorithms on Matrix Manifolds
Absil, P.-A., Mahony, R., & Sepulchre, R. (2008) · Princeton University Press eBooks · doi:10.1515/9781400830244 · Cited by 1,840
Multilayer formulation of the multiconfiguration time-dependent Hartree theory
Wang, H., & Thoss, M. (2003) · The Journal of Chemical Physics · doi:10.1063/1.1580111 · Cited by 932 · cited 3×
“5) were already used in the quantum chemists community [32]; see also [20, p.”
BackgroundMethodTime-Dependent Variational Principle for Quantum Lattices
Haegeman, J., Cirac, J. I., Osborne, T. J., Pižorn, I., Verschelde, H., & Verstraete, F. (2011) · Physical Review Letters · doi:10.1103/physrevlett.107.070601 · Cited by 865 · cited 1×
“In particular, the geometric approach underlying this work has lately been used for the simulation of realand imaginary-time dynamics for infinite one-dimensional quantum lattices by MPS [10].”
MethodTensor Spaces and Numerical Tensor Calculus
Hackbusch, W. (2012) · Springer Series in Computational Mathematics · doi:10.1007/978-3-642-28027-6 · Cited by 607
Hierarchical Singular Value Decomposition of Tensors
Grasedyck, L. (2010) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/090764189 · Cited by 508
Breaking the Curse of Dimensionality, Or How to Use SVD in Many Dimensions
Oseledets, I. V., & Tyrtyshnikov, E. E. (2009) · SIAM Journal on Scientific Computing · doi:10.1137/090748330 · Cited by 459 · cited 2×
“, by Grasedyck [7], or a front-to-back splitting which (with minor adaptations, see Remark 1 below) leads to the TT format due to Oseledets and Tyrtyshnikov [25, 26].”
A New Scheme for the Tensor Representation
Hackbusch, W., & Kühn, S. (2009) · Journal of Fourier Analysis and Applications · doi:10.1007/s00041-009-9094-9 · Cited by 453 · cited 3×
“Aside from some necessary concepts from [7, 9], we also introduce notation similar to that of [19, 18, 30] that is better suited to our analysis.”
MethodResultFrom Quantum to Classical Molecular Dynamics: Reduced Models and Numerical Analysis
Lubich, C. (2008) · Zurich Lectures in Advanced Mathematics · doi:10.4171/067 · Cited by 366 · cited 2×
“In the context of the time-dependent Schrödinger equation, the approach is known as the Dirac–Frenkel time-dependent variational principle [20].”
BackgroundMethodSelf-citationThe Alternating Linear Scheme for Tensor Optimization in the Tensor Train Format
Holtz, S., Rohwedder, T., & Schneider, R. (2012) · SIAM Journal on Scientific Computing · doi:10.1137/100818893 · Cited by 337 · cited 2×
“Approximately computing the dynamical approximation to A+ tΔA with one or few Euler steps provides an alternative to truncation steps by higher order SVD [7], or by local optimization methods like alternating least squares methods [11].”
MethodSelf-citationDynamical Low‐Rank Approximation
Koch, O., & Lubich, C. (2007) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/050639703 · Cited by 301 · cited 6×
“The proofs are the same as for the corresponding results in [15] and [16], using Lemma 4.”
BackgroundMethodResultInfluentialSelf-citationSimulating strongly correlated quantum systems with tree tensor networks
Murg, V., Verstraete, F., Legeza, Ö., & Noack, R. M. (2010) · Physical Review B · doi:10.1103/physrevb.82.205105 · Cited by 245 · cited 1×
“Formats of this type have been used successfully in the quantum physics and chemistry communities in the last decade [21, 23, 27, 32].”
MethodA projection method to solve linear systems in tensor format
Ballani, J., & Grasedyck, L. (2013) · Numerical Linear Algebra with Applications · doi:10.1002/nla.1818 · Cited by 203
On manifolds of tensors of fixed TT-rank
Holtz, S., Rohwedder, T., & Schneider, R. (2012) · Numerische Mathematik · doi:10.1007/s00211-011-0419-7 · Cited by 186 · cited 3×
“Analogous to the Tucker format, both HT and TT allow the definition of a rank vector k (to be specified in section 2), and the manifold of tensors of fixed HT or TT rank can be shown to be an embedded manifold [12, 30], to which the above dynamical tensor approximation approach may then be applied.”
MethodSelf-citationKolda, T. G., & Bader, B. W. (2006) · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · Cited by 154
Dynamical Tensor Approximation
Koch, O., & Lubich, C. (2010) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/09076578x · Cited by 131 · cited 6×
“The proofs are the same as for the corresponding results in [15] and [16], using Lemma 4.”
BackgroundMethodResultInfluentialSelf-citationThe geometry of algorithms using hierarchical tensors
Uschmajew, A., & Vandereycken, B. (2013) · Linear Algebra and its Applications · doi:10.1016/j.laa.2013.03.016 · Cited by 111
Self-citationA Riemannian Optimization Approach for Computing Low-Rank Solutions of Lyapunov Equations
Vandereycken, B., & Vandewalle, S. (2010) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/090764566 · Cited by 108 · cited 1×
“The dynamical approximation may thus be used as a retraction as required in optimization methods on low-rank manifolds like CG and Newton; see [1, 31].”
MethodSelf-citationComplete-graph tensor network states: A new fermionic wave function ansatz for molecules
Marti, K. H., Bauer, B., Reiher, M., Troyer, M., & Verstraete, F. (2010) · doi:10.3929/ethz-b-000159263 · Cited by 89
Complete-graph tensor network states: a new fermionic wave function ansatz for molecules
Marti, K. H., Bauer, B., Reiher, M., Troyer, M., & Verstraete, F. (2010) · New Journal of Physics · doi:10.1088/1367-2630/12/10/103008 · Cited by 87 · cited 1×
“Formats of this type have been used successfully in the quantum physics and chemistry communities in the last decade [21, 23, 27, 32].”
MethodPreconditioned Low-Rank Methods for High-Dimensional Elliptic PDE Eigenvalue Problems
Kreßner, D., & Tobler, C. (2011) · Computational Methods in Applied Mathematics · doi:10.2478/cmam-2011-0020 · Cited by 84 · cited 1×
“Aside from some necessary concepts from [7, 9], we also introduce notation similar to that of [19, 18, 30] that is better suited to our analysis.”
ResultDynamical low-rank approximation: applications and numerical experiments
Nonnenmacher, A., & Lubich, C. (2008) · Mathematics and Computers in Simulation · doi:10.1016/j.matcom.2008.03.007 · Cited by 56 · cited 1×
“, kd) in [15, 24] and [16], respectively.”
Self-citationAn error analysis of the multi-configuration time-dependent Hartree method of quantum dynamics
Conte, D., & Lubich, C. (2010) · ESAIM Mathematical Modelling and Numerical Analysis · doi:10.1051/m2an/2010018 · Cited by 48
Self-citationhtucker – a matlab toolbox for tensors in hierarchical Tucker format
Kreßner, D., & Tobler, C. (2012) · Cited by 37
Efficient time-stepping scheme for dynamics on TT-manifolds
Khoromskij, B. N., Oseledets, I. V., & Schneider, R. (2012) · Cited by 20
Self-citationDolgov, S., Khoromskij, B. N., & Oseledets, I. V. (2012) · Cited by 14
Research works printed but not linked: 1
- G. W. Stewart, Matrix Algorithms: Vol. $1$, Basic Decompositions , SIAM, Philadelphia, 1998. · DOI
Cited by
Unifying time evolution and optimization with matrix product states
Haegeman, J., Lubich, C., Oseledets, I., Vandereycken, B., & Verstraete, F. (2016) · Physical Review B · doi:10.1103/physrevb.94.165116 · Cited by 804
Self-citationA literature survey of low‐rank tensor approximation techniques
Grasedyck, L., Kressner, D., & Tobler, C. (2013) · GAMM-Mitteilungen · doi:10.1002/gamm.201310004 · Cited by 577
“Initially proposed for low-rank matrix manifolds in [166], dynamical low-rank methods have been extended to low-rank tensors in Tucker [167, 199], TT/MPS [118, 123, 163, 186], and HT [9, 186, 255] decomposition.”
Tensor product methods and entanglement optimization for ab initio quantum chemistry
Szalay, S., Pfeffer, M., Murg, V., Barcza, G., Verstraete, F., Schneider, R., & Legeza, Ö. (2015) · International Journal of Quantum Chemistry · doi:10.1002/qua.24898 · Cited by 314
Self-citationAlternating Minimal Energy Methods for Linear Systems in Higher Dimensions
Dolgov, S. V., & Savostyanov, D. V. (2014) · SIAM Journal on Scientific Computing · doi:10.1137/140953289 · Cited by 253
“Several theoretical issues were addressed in [43], and the numerical splitting w.”
Cichocki, A., Lee, N., Oseledets, I., Phan, A.-H., Zhao, Q., Sugiyama, M., & Mandic, D. P. (2017) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000067 · Cited by 224
Metadata
- Authors
- Lubich, C., Rohwedder, T., Schneider, R., & Vandereycken, B.
- Venue
- SIAM Journal on Matrix Analysis and Applications
- Published
- 2013-01
- Volume
- 34
- Issue
- 2
- Pages
- 470-494
- ISSN
- 0895-4798 (ISSN-L)
- 1095-7162
- OpenAlex
- W2067067855
- Cited by
- 163
- References
- 31
- Citations by year
- 2026: 8 · 2025: 8 · 2024: 17 · 2023: 14 · 2022: 16 · 2021: 14 · 2020: 13 · 2019: 13 · 2018: 9 · 2017: 7 · 2016: 9 · 2015: 12 · 2014: 13 · 2013: 8
- FWCI
- 4.36
- Type
- journal article
- Language
- en
- Open access
- green
- infoscience.epfl.ch
- Locations
- SIAM Journal on Matrix Analysis and Applications (publishedVersion)
- Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion)
- Infoscience (Ecole Polytechnique Fédérale de Lausanne) (submittedVersion, cc-by-nc-nd)
- Topics
- Physical Sciences › Mathematics › Computational Mathematics › Tensor decomposition and applications
- Physical Sciences › Mathematics › Numerical Analysis › Numerical Solvers for Ordinary Differential Equations
- Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Model Reduction and Neural Networks
- Keywords
- tensor train format, dynamical low-rank approximation, curvature estimates, optimization algorithms
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.