Approximation theory of tree tensor networks: tensorized multivariate functions
Abstract
We study the approximation of multivariate functions with tensor networks (TNs), providing some answers to the following two questions: “ what are the approximation capabilities of TNs for functions from classical smoothness classes? ” and “ what are the properties of the class of functions that can be approximated with TNs with a certain performance? ” As a partial answer to the former, we show that TNs can (near to) optimally replicate h h -uniform and h h -adaptive spline approximation, for any smoothness order of the target function. TNs thus exhibit universal expressivity w.r.t. isotropic, anisotropic and mixed smoothness spaces that is comparable with more general neural networks families such as deep rectified linear unit networks. Put differently, TNs have the capacity to (near to) optimally approximate many function classes – without being adapted to the particular class in question. As a partial answer to the latter, as a candidate model class we consider approximation classes of TNs and show that these are (quasi-)Banach spaces, that many types of classical smoothness spaces are continuously embedded into said approximation classes and that TNs approximation classes are themselves not embedded in any classical smoothness space. In other words, TNs can efficiently approximate functions that lie beyond classical smoothness spaces.
Related
- Compressing multivariate functions with tree tensor networks
- Deep ReLU Networks Can Fail the Best Approximation Property for Continuous Target Functions
- Optimal approximation of smooth functions by ReLU neural networks
- Backward log orthogonal functions and their approximation theory
- MIMO and Multi-Group Comparison Problem in Process Control: A Multivariate Statistical Framework for Systems Represented with Frequency Response Functions
BibTeX
@article{ali2026approximation,
title = {{Approximation theory of tree tensor networks: tensorized multivariate functions}},
author = {Mazen Ali and Anthony Nouy},
journal = {Mathematics of Computation},
year = {2026},
month = oct,
doi = {10.1090/mcom/4257},
url = {https://doi.org/10.1090/mcom/4257},
note = {CC BY},
}
RIS
TY - JOUR
TI - Approximation theory of tree tensor networks: tensorized multivariate functions
AU - Mazen Ali
AU - Anthony Nouy
PY - 2026
DA - 2026/10/05
JO - Mathematics of Computation
DO - 10.1090/mcom/4257
UR - https://doi.org/10.1090/mcom/4257
AB - We study the approximation of multivariate functions with tensor networks (TNs), providing some answers to the following two questions: “ what are the approximation capabilities of TNs for functions from classical smoothness classes? ” and “ what are the properties of the class of functions that can be approximated with TNs with a certain performance? ” As a partial answer to the former, we show that TNs can (near to) optimally replicate h h -uniform and h h -adaptive spline approximation, for any smoothness order of the target function. TNs thus exhibit universal expressivity w.r.t. isotropic, anisotropic and mixed smoothness spaces that is comparable with more general neural networks families such as deep rectified linear unit networks. Put differently, TNs have the capacity to (near to) optimally approximate many function classes – without being adapted to the particular class in question. As a partial answer to the latter, as a candidate model class we consider approximation classes of TNs and show that these are (quasi-)Banach spaces, that many types of classical smoothness spaces are continuously embedded into said approximation classes and that TNs approximation classes are themselves not embedded in any classical smoothness space. In other words, TNs can efficiently approximate functions that lie beyond classical smoothness spaces.
SN - 0025-5718
SN - 1088-6842
ER -