Estimating the error variance in a high-dimensional linear model
arXiv:1712.02412 · doi:10.1093/biomet/asz017
Abstract
The lasso has been studied extensively as a tool for estimating the coefficient vector in the high-dimensional linear model; however, considerably less is known about estimating the error variance in this context. In this paper, we propose the natural lasso estimator for the error variance, which maximizes a penalized likelihood objective. A key aspect of the natural lasso is that the likelihood is expressed in terms of the natural parameterization of the multiparameter exponential family of a Gaussian with unknown mean and variance. The result is a remarkably simple estimator of the error variance with provably good performance in terms of mean squared error. These theoretical results do not require placing any assumptions on the design matrix or the true regression coefficients. We also propose a companion estimator, called the organic lasso, which theoretically does not require tuning of the regularization parameter. Both estimators do well empirically compared to preexisting methods, especially in settings where successful recovery of the true support of the coefficient vector is hard. Finally, we show that existing methods can do well under fewer assumptions than previously known, thus providing a fuller story about the problem of estimating the error variance in high-dimensional linear models.
arXiv commentBiometrika(2019)
References in corpus (13)
- Pathwise coordinate optimization
- On asymptotically optimal confidence regions and tests for high-dimensional models
- On the conditions used to prove oracle results for the Lasso
- Confidence Intervals and Hypothesis Testing for High-Dimensional Regression
- Square-Root Lasso: Pivotal Recovery of Sparse Signals via Conic Programming
- A significance test for the lasso
- Exact post-selection inference, with application to the lasso
- Statistical significance in high-dimensional linear models
- L1-Penalization for Mixture Regression Models
- On the Prediction Performance of the Lasso
- Sparse Matrix Inversion with Scaled Lasso
- Prediction error of cross-validated Lasso
- The Simulator: An Engine to Streamline Simulations
Cited by in corpus (5)
- Learning High-dimensional Gaussian Graphical Models under Total Positivity without Adjustment of Tuning Parameters
- yaglm: a Python package for fitting and tuning generalized linear models that supports structured, adaptive and non-convex penalties
- Generalized Matrix Decomposition Regression: Estimation and Inference for Two-way Structured Data
- Learning Gaussian DAGs from Network Data
- Scalable simultaneous inference in high-dimensional linear regression models