A Riemannian Trust Region Method for the Canonical Tensor Rank Approximation Problem
Authors:
- Paul BreidingPBPaul Breiding
- Nick VannieuwenhovenNVNick Vannieuwenhoven
- 1 January 2018
- journal article
- research article
- Published by Society for Industrial & Applied Mathematics (SIAM) in SIAM Journal on Optimization
- Vol. 28 (3) , 2435-2465
Abstract
The canonical tensor rank approximation problem (TAP) consists of approximating a real-valued tensor by one of low canonical rank, which is a challenging nonlinear, nonconvex, constrained optimization problem, where the constraint set forms a nonsmooth semialgebraic set. We introduce a Riemannian Gauss--Newton method with trust region for solving small-scale, dense TAPs. The novelty of our approach is threefold. First, we parametrize the constraint set as the Cartesian product of Segre manifolds, hereby formulating the TAP as a Riemannian optimization problem, and we argue why this parametrization is theoretically a good choice. Second, an original sequentially truncated higher-order singular value decomposition (ST-HOSVD) based retraction operator is proposed. Third, we introduce a hot restart mechanism that efficiently detects when the optimization process is tending to an ill-conditioned tensor rank decomposition and that often yields a quick escape path from such spurious decompositions. Numerical exp...Keywords
All Related Versions
Funding Information
- Deutsche Forschungsgemeinschaft (BU 1371/2-2)
- Fonds Wetenschappelijk Onderzoek
This publication has 58 references indexed in Scilit:
- Scalable tensor factorizations for incomplete dataChemometrics and Intelligent Laboratory Systems, 2011
- First and second-order derivatives for CP and INDSCALChemometrics and Intelligent Laboratory Systems, 2011
- Identifiability of parameters in latent structure models with many observed variablesThe Annals of Statistics, 2009
- Fluorescence analysis of dissolved organic matter in natural, waste and polluted waters—a reviewRiver Research and Applications, 2007
- Cubic regularization of Newton method and its global performanceMathematical Programming, 2006
- Algebraic geometry of Bayesian networksJournal of Symbolic Computation, 2005
- Newton's method on Riemannian manifolds and a geometric model for the human spineIMA Journal of Numerical Analysis, 2002
- The N-way Toolbox for MATLABChemometrics and Intelligent Laboratory Systems, 2000
- A New Algorithm to Solve Parafac-ModelBehaviormetrika, 1982
- Strategies for analyzing data from video fluorometric monitoring of liquid chromatographic effluentsAnalytical Chemistry, 1981