DeepAI AI Chat
Log In Sign Up

On Tensor Train Rank Minimization: Statistical Efficiency and Scalable Algorithm

by   Masaaki Imaizumi, et al.

Tensor train (TT) decomposition provides a space-efficient representation for higher-order tensors. Despite its advantage, we face two crucial limitations when we apply the TT decomposition to machine learning problems: the lack of statistical theory and of scalable algorithms. In this paper, we address the limitations. First, we introduce a convex relaxation of the TT decomposition problem and derive its error bound for the tensor completion task. Next, we develop an alternating optimization method with a randomization technique, in which the time complexity is as efficient as the space complexity is. In experiments, we numerically confirm the derived bounds and empirically demonstrate the performance of our method with a real higher-order tensor.


page 1

page 2

page 3

page 4


High-order Tensor Completion for Data Recovery via Sparse Tensor-train Optimization

In this paper, we aim at the problem of tensor data completion. Tensor-t...

Tensor train rank minimization with nonlocal self-similarity for tensor completion

The tensor train (TT) rank has received increasing attention in tensor c...

Scalable symmetric Tucker tensor decomposition

We study the best low-rank Tucker decomposition of symmetric tensors, ad...

Incoherent Tensor Norms and Their Applications in Higher Order Tensor Completion

In this paper, we investigate the sample size requirement for a general ...

Score Function Features for Discriminative Learning: Matrix and Tensor Framework

Feature learning forms the cornerstone for tackling challenging learning...

Score Function Features for Discriminative Learning

Feature learning forms the cornerstone for tackling challenging learning...

Normal Forms for Tensor Rank Decomposition

We propose a new algorithm for computing the tensor rank decomposition o...