Matrix Reordering for Noisy Disordered Matrices: Optimality and Computationally Efficient Algorithms
Abstract
Motivated by applications in single-cell biology and metagenomics, we investigate the problem of matrix reordering based on a noisy disordered monotone Toeplitz matrix model. We establish the fundamental statistical limit for this problem in a decision-theoretic framework and demonstrate that a constrained least squares estimator achieves the optimal rate. However, due to its computational complexity, we analyze a popular polynomial-time algorithm, spectral seriation, and show that it is suboptimal. To address this, we propose a novel polynomial-time adaptive sorting algorithm with guaranteed performance improvement. Simulations and analyses of two real single-cell RNA sequencing datasets demonstrate the superiority of our algorithm over existing methods.
Cite
@article{arxiv.2201.06438,
title = {Matrix Reordering for Noisy Disordered Matrices: Optimality and Computationally Efficient Algorithms},
author = {T. Tony Cai and Rong Ma},
journal= {arXiv preprint arXiv:2201.06438},
year = {2023}
}
Comments
accepted by IEEE Transactions on Information Theory