Group-Invariant Tensor Train Networks for Supervised Learning

DOI10.1137/22M1506857MaRDI QIDQ6070296zbMATH OpenOpenAlexFDO

Authors Nick Vannieuwenhoven

Publication date 20 November 2023

Published in SIAM Journal on Mathematics of Data Science (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2206.15051

zbMATH Keywords

group equivariancegroup invariancerepresentation theorysupervised learningtensor networkstensor trains

Mathematics Subject Classification ID

Linear transformations, semilinear transformations (15A04) Multilinear algebra, tensor calculus (15A69) Representations of finite symmetric groups (20C30) Numerical computation of eigenvalues and eigenvectors of matrices (65F15) Learning and adaptive systems in artificial intelligence (68T05) Computational aspects of data analysis and big data (68T09)

Abstract: Invariance has recently proven to be a powerful inductive bias in machine learning models. One such class of predictive or generative models are tensor networks. We introduce a new numerical algorithm to construct a basis of tensors that are invariant under the action of normal matrix representations of an arbitrary discrete group. This method can be up to several orders of magnitude faster than previous approaches. The group-invariant tensors are then combined into a group-invariant tensor train network, which can be used as a supervised machine learning model. We applied this model to a protein binding classification problem, taking into account problem-specific invariances, and obtained prediction accuracy in line with state-of-the-art deep learning approaches.

Recommendations

• Probabilistic symmetries and invariant neural networks
• A unifying framework for invariant pattern recognition
• Accelerated discovery of machine-learned symmetries: deriving the exceptional Lie groups \(G_2\), \(F_4\) and \(E_6\)
• Representation theory and invariant neural networks

Cites work

• A new scheme for the tensor representation
• A randomized algorithm for testing nonsingularity of structured matrices with an application to asserting nondefectivity of Segre varieties
• Algorithm 862
• Desingularization of bounded-rank matrix sets
• Generalized Eigenvalue Problems for Rectangular Matrices
• Hand-waving and interpretive dance: an introductory course on tensor networks
• Hierarchical Singular Value Decomposition of Tensors
• scientific article; zbMATH DE number 1703931 (Why is no real title available?)
• scientific article; zbMATH DE number 3650737 (Why is no real title available?)
• scientific article; zbMATH DE number 51906 (Why is no real title available?)
• scientific article; zbMATH DE number 194139 (Why is no real title available?)
• Multilinear algebra. 2nd ed
• Tensor-train decomposition
• The density-matrix renormalization group in the age of matrix product states

Cited in

(1)

• Accelerated discovery of machine-learned symmetries: deriving the exceptional Lie groups \(G_2\), \(F_4\) and \(E_6\)

This page was built for publication: Group-Invariant Tensor Train Networks for Supervised Learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6070296)