Near-optimal sample complexity for convex tensor completion
Abstract
We analyze low rank tensor completion (TC) using noisy measurements of a subset of the tensor. Assuming a rank-, order-, tensor where , the best sampling complexity that was achieved is , which is obtained by solving a tensor nuclear-norm minimization problem. However, this bound is significantly larger than the number of free variables in a low rank tensor which is . In this paper, we show that by using an atomic-norm whose atoms are rank- sign tensors, one can obtain a sample complexity of . Moreover, we generalize the matrix max-norm definition to tensors, which results in a max-quasi-norm (max-qnorm) whose unit ball has small Rademacher complexity. We prove that solving a constrained least squares estimation using either the convex atomic-norm or the nonconvex max-qnorm results in optimal sample complexity for the problem of low-rank tensor completion. Furthermore, we show that these bounds are nearly minimax rate-optimal. We also provide promising numerical results for max-qnorm constrained tensor completion, showing improved recovery results compared to matricization and alternating least squares.
Cite
@article{arxiv.1711.04965,
title = {Near-optimal sample complexity for convex tensor completion},
author = {Navid Ghadermarzy and Yaniv Plan and Özgür Yılmaz},
journal= {arXiv preprint arXiv:1711.04965},
year = {2017}
}