Abstract
This paper is concerned with the approximation of tensors using tree-based tensor formats, which are tensor networks whose graphs are dimension partition trees. We consider Hilbert tensor spaces of multivariate functions defined on a product set equipped with a probability measure. This includes the case of multidimensional arrays corresponding to finite product sets. We propose and analyse an algorithm for the construction of an approximation using only point evaluations of a multivariate function, or evaluations of some entries of a multidimensional array. The algorithm is a variant of higher-order singular value decomposition which constructs a hierarchy of subspaces associated with the different nodes of the tree and a corresponding hierarchy of interpolation operators. Optimal subspaces are estimated using empirical principal component analysis of interpolations of partial random evaluations of the function. The algorithm is able to provide an approximation in any tree-based format with either a prescribed rank or a prescribed relative error, with a number of evaluations of the order of the storage complexity of the approximation format. Under some assumptions on the estimation of principal components, we prove that the algorithm provides either a quasi-optimal approximation with a given rank, or an approximation satisfying the prescribed relative error, up to constants depending on the tree and the properties of interpolation operators. The analysis takes into account the discretization errors for the approximation of infinite-dimensional tensors. For a tensor with finite and known rank in a tree-based format, the algorithm is able to recover the tensor in a stable way using a number of evaluations equal to the storage complexity of the representation of the tensor in this format. Several numerical examples illustrate the main results and the behavior of the algorithm for the approximation of high-dimensional functions using hierarchical Tucker or tensor train tensor formats, and the approximation of univariate functions using tensorization.
Similar content being viewed by others
Notes
Note that \(V'\subset \tilde{U}'\) and we may have \(W \not \subset V'.\)
Note that since \({\mathrm {rank}}_{\{d\}}(v) = {\mathrm {rank}}_{\{1,\ldots ,d-1\}}(v)\), adding the node \(\{d\}\) in the set of active nodes A would yield an equivalent tensor format.
For all \(m \ge r_\alpha \), we have \(\mathcal {P}_{U_\alpha ^\star } u_{m} = \sum _{k=1}^{m} \sigma _k^\alpha (P_{U_\alpha ^\star } u_k^\alpha ) \otimes u_k^{\alpha ^c} = \sum _{k=1}^{r_\alpha } \sigma _k^\alpha u_k^\alpha \otimes u_k^{\alpha ^c} = u_{r_\alpha }\). Then using the continuity of \(\mathcal {P}_{U_\alpha ^\star }\) and taking the limit with m, we obtain \(\mathcal {P}_{U_\alpha ^\star } u = u_{r_\alpha }\).
Note that when \(\mathcal {H}\) is equipped with a norm stronger than the norm in \(L^2_{\mu }(\mathcal {X})\), then \(\Vert \cdot \Vert _\alpha \) does not coincides with the norm \(\Vert \cdot \Vert \) on \(\mathcal {H}\), so that the subspaces solutions of (7) and (11) will be different in general.
For the last example, \(\mathcal {X}\) is a finite product set equipped with the uniform measure and \(L^2_\mu (\mathcal {X})\) then corresponds to the space of multidimensional arrays equipped with the canonical norm.
References
Bachmayr, M., Schneider, R., Uschmajew, A.: Tensor networks and hierarchical tensors for the solution of high-dimensional partial differential equations. Found. Comput. Math. 16(6), 1423–1472 (2016)
Ballani, J., Grasedyck, L., Kluge, M.: Black box approximation of tensors in hierarchical Tucker format. Linear Algebra Appl. 438(2), 639–657 (2013). Tensors and Multilinear Algebra
Blanchard, G., Bousquet, O., Zwald, L.: Statistical properties of kernel principal component analysis. Mach. Learn. 66(2–3), 259–294 (2007)
Bungartz, H.-J., Griebel, M.: Sparse grids. Acta Numer. 13, 147–269 (2004)
Chevreuil, M., Lebrun, R., Nouy, A., Rai, P.: A least-squares method for sparse low rank approximation of multivariate functions. SIAM/ASA J. Uncertain. Quantif. 3(1), 897–921 (2015)
Cohen, A., DeVore, R.: Approximation of high-dimensional parametric pdes. Acta Numer. 24, 1–159 (2015)
Cohen, N., Sharir, O., Shashua, A.: On the expressive power of deep learning: a tensor analysis. In: Conference on Learning Theory, pp. 698–728 (2016)
De Lathauwer, L., De Moor, B., Vandewalle, J.: A multilinear singular value decomposition. SIAM J. Matrix Anal. Appl. 21(4), 1253–1278 (2000)
de Silva, V., Lim, L.-H.: Tensor rank and ill-posedness of the best low-rank approximation problem. SIAM J. Matrix Anal. Appl. 30(3), 1084–1127 (2008)
DeVore, R.A.: Nonlinear approximation. Acta Numer. 7, 51–150 (1998)
Doostan, A., Validi, A., Iaccarino, G.: Non-intrusive low-rank separated approximation of high-dimensional stochastic models. Comput. Methods Appl. Mech. Eng. 263, 42–55 (2013)
Espig, M., Grasedyck, L., Hackbusch, W.: Black box low tensor-rank approximation using fiber-crosses. Constr. Approx. 30, 557–597 (2009)
Falcó, A., Hackbusch, W.: On minimal subspaces in tensor representations. Found. Comput. Math. 12, 765–803 (2012)
Falco, A., Hackbusch, W., Nouy, A.: Geometric Structures in Tensor Representations (Final Release). ArXiv e-prints (2015)
Falcó, A., Hackbusch, W., Nouy, A.: On the Dirac–Frenkel variational principle on tensor Banach spaces. Found. Comput. Math. (2018). https://doi.org/10.1007/s10208-018-9381-4
Falcó, A., Hackbusch, W., Nouy, A.: Tree-based tensor formats. SeMA J. (2018). https://doi.org/10.1007/s40324-018-0177-x
Grasedyck, L.: Hierarchical singular value decomposition of tensors. SIAM J. Matrix Anal. Appl. 31, 2029–2054 (2010)
Grasedyck, L., Kressner, D., Tobler, C.: A literature survey of low-rank tensor approximation techniques. GAMM-Mitteilungen 36(1), 53–78 (2013)
Grelier, E., Nouy, A., Chevreuil, M.: Learning with tree-based tensor formats (2018). arXiv e-prints arXiv:1811.04455
Hackbusch, W.: Tensor Spaces and Numerical Tensor Calculus, Volume 42 of Springer Series in Computational Mathematics. Springer, Heidelberg (2012)
Hackbusch, W., Kuhn, S.: A new scheme for the tensor representation. J. Fourier Anal. Appl. 15(5), 706–722 (2009)
Hillar, C., Lim, L.-H.: Most tensor problems are np-hard. J. ACM (JACM) 60(6), 45 (2013)
Holtz, S., Rohwedder, T., Schneider, R.: On manifolds of tensors of fixed tt-rank. Numer. Math. 120(4), 701–731 (2012)
Jirak, M., Wahl, M.: A tight \(\sin \varTheta \) theorem for empirical covariance operators. ArXiv e-prints (2018)
Jirak, M., Wahl, M.: Relative perturbation bounds with applications to empirical covariance operators. ArXiv e-prints (2018)
Khoromskij, B.: O (dlog n)-quantics approximation of nd tensors in high-dimensional numerical modeling. Constr. Approx. 34(2), 257–280 (2011)
Khoromskij, B.: Tensors-structured numerical methods in scientific computing: survey on recent advances. Chemometr. Intell. Lab. Syst. 110(1), 1–19 (2012)
Kolda, T.G., Bader, B.W.: Tensor decompositions and applications. SIAM Rev. 51(3), 455–500 (2009)
Kressner, D., Steinlechner, M., Uschmajew, A.: Low-rank tensor methods with subspace correction for symmetric eigenvalue problems. SIAM J. Sci. Comput. 36(5), A2346–A2368 (2014)
Lubich, C., Rohwedder, T., Schneider, R., Vandereycken, B.: Dynamical approximation by hierarchical tucker and tensor-train tensors. SIAM J. Matrix Anal. Appl. 34(2), 470–494 (2013)
Luu, T.H., Maday, Y., Guillo, M., Guérin, P.: A new method for reconstruction of cross-sections using Tucker decomposition. J. Comput. Phys. 345, 189–206 (2017)
Maday, Y., Nguyen, N.C., Patera, A.T., Pau, G.S.H.: A general multipurpose interpolation procedure: the magic points. Commun. Pure Appl. Anal. 8(1), 383–404 (2009)
Megginson, R.E: An Introduction to Banach Space Theory, Vol. 183. Springer, Berlin (2012)
Nouy, A.: Low-rank methods for high-dimensional approximation and model order reduction. In: Benner, P., Cohen, A., Ohlberger, M., Willcox, K. (eds.) Model Reduction and Approximation: Theory and Algorithms. SIAM, Philadelphia (2017)
Nouy, A.: Low-Rank Tensor Methods for Model Order Reduction, pp. 857–882. Springer, Cham (2017)
Orus, R.: A practical introduction to tensor networks: matrix product states and projected entangled pair states. Ann. Phys. 349, 117–158 (2014)
Oseledets, I.: Tensor-train decomposition. SIAM J. Sci. Comput. 33(5), 2295–2317 (2011)
Oseledets, I., Tyrtyshnikov, E.: Breaking the curse of dimensionality, or how to use svd in many dimensions. SIAM J. Sci. Comput. 31(5), 3744–3759 (2009)
Oseledets, I., Tyrtyshnikov, E.: TT-cross approximation for multidimensional arrays. Linear Algebra Appl. 432(1), 70–88 (2010)
Oseledets, I., Tyrtyshnikov, E.: Algebraic wavelet transform via quantics tensor train decomposition. SIAM J. Sci. Comput. 33(3), 1315–1328 (2011)
Reiß, M., Wahl, M.: Non-asymptotic upper bounds for the reconstruction error of PCA (2016). arXiv preprint arXiv:1609.03779
Schneider, R., Uschmajew, A.: Approximation rates for the hierarchical tensor format in periodic Sobolev spaces. J. Complex. 30(2), 56–71 (2014). Dagstuhl 2012
Temlyakov, V.: Nonlinear methods of approximation. Found. Comput. Math. 3(1), 33–107 (2003)
Temlyakov, V.: Greedy Approximation. Cambridge Monographs on Applied and Computational Mathematics. Cambridge University Press, Cambridge (2011)
Uschmajew, A., Vandereycken, B.: The geometry of algorithms using hierarchical tensors. Linear Algebra Appl. 439(1), 133–166 (2013)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Nouy, A. Higher-order principal component analysis for the approximation of tensors in tree-based low-rank formats. Numer. Math. 141, 743–789 (2019). https://doi.org/10.1007/s00211-018-1017-8
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00211-018-1017-8