English

Regularization Methods Based on the $L_q$-Likelihood for Linear Models with Heavy-Tailed Errors

Methodology 2020-10-28 v1

Abstract

We propose regularization methods for linear models based on the LqL_q-likelihood, which is a generalization of the log-likelihood using a power function. Some heavy-tailed distributions are known as qq-normal distributions. We find that the proposed methods for linear models with qq-normal errors coincide with the regularization methods that are applied to the normal linear model. The proposed methods work well and efficiently, and can be computed using existing packages. We examine the proposed methods using numerical experiments, showing that the methods perform well, even when the error is heavy-tailed.

Keywords

Cite

@article{arxiv.2008.10876,
  title  = {Regularization Methods Based on the $L_q$-Likelihood for Linear Models with Heavy-Tailed Errors},
  author = {Yoshihiro Hirose},
  journal= {arXiv preprint arXiv:2008.10876},
  year   = {2020}
}
R2 v1 2026-06-23T18:05:05.658Z