English

Element-wise estimation error of a total variation regularized estimator for change point detection

Statistics Theory 2019-01-07 v1 Statistics Theory

Abstract

This work studies the total variation regularized 2\ell_2 estimator (fused lasso) in the setting of a change point detection problem. Compared with existing works that focus on the sum of squared estimation errors, we give bound on the element-wise estimation error. Our bound is nearly optimal in the sense that the sum of squared error matches the best existing result, up to a logarithmic factor. This analysis of the element-wise estimation error allows a screening method that can approximately detect all the change points. We also generalize this method to the muitivariate setting, i.e., to the problem of group fused lasso.

Keywords

Cite

@article{arxiv.1901.00914,
  title  = {Element-wise estimation error of a total variation regularized estimator for change point detection},
  author = {Teng Zhang},
  journal= {arXiv preprint arXiv:1901.00914},
  year   = {2019}
}
R2 v1 2026-06-23T07:02:40.640Z