English

Bootstrap-Based Regularization for Low-Rank Matrix Estimation

Methodology 2016-06-29 v3 Machine Learning Machine Learning

Abstract

We develop a flexible framework for low-rank matrix estimation that allows us to transform noise models into regularization schemes via a simple bootstrap algorithm. Effectively, our procedure seeks an autoencoding basis for the observed matrix that is stable with respect to the specified noise model; we call the resulting procedure a stable autoencoder. In the simplest case, with an isotropic noise model, our method is equivalent to a classical singular value shrinkage estimator. For non-isotropic noise models, e.g., Poisson noise, the method does not reduce to singular value shrinkage, and instead yields new estimators that perform well in experiments. Moreover, by iterating our stable autoencoding scheme, we can automatically generate low-rank estimates without specifying the target rank as a tuning parameter.

Keywords

Cite

@article{arxiv.1410.8275,
  title  = {Bootstrap-Based Regularization for Low-Rank Matrix Estimation},
  author = {Julie Josse and Stefan Wager},
  journal= {arXiv preprint arXiv:1410.8275},
  year   = {2016}
}

Comments

To appear in the Journal of Machine Learning Research

R2 v1 2026-06-22T06:41:29.575Z