A statistical mechanics approach to de-biasing and uncertainty estimation in Lasso for random measurements
From MaRDI portal
Publication:3303301
DOI10.1088/1742-5468/AACE2EzbMATH Open1456.62151arXiv1803.09927OpenAlexW3100697818MaRDI QIDQ3303301FDOQ3303301
Authors: Takashi Takahashi, Yoshiyuki Kabashima 
Publication date: 11 August 2020
Published in: Journal of Statistical Mechanics: Theory and Experiment (Search for Journal in Brave)
Abstract: In high-dimensional statistical inference in which the number of parameters to be estimated is larger than that of the holding data, regularized linear estimation techniques are widely used. These techniques have, however, some drawbacks. First, estimators are biased in the sense that their absolute values are shrunk toward zero because of the regularization effect. Second, their statistical properties are difficult to characterize as they are given as numerical solutions to certain optimization problems. In this manuscript, we tackle such problems concerning LASSO, which is a widely used method for sparse linear estimation, when the measurement matrix is regarded as a sample from a rotationally invariant ensemble. We develop a new computationally feasible scheme to construct a de-biased estimator with a confidence interval and conduct hypothesis testing for the null hypothesis that a certain parameter vanishes. It is numerically confirmed that the proposed method successfully de-biases the LASSO estimator and constructs confidence intervals and p-values by experiments for noisy linear measurements.
Full work available at URL: https://arxiv.org/abs/1803.09927
Recommendations
- Debiasing the Lasso: optimal sample size for Gaussian designs
- Debiasing the debiased Lasso with bootstrap
- Confidence Intervals and Hypothesis Testing for High-Dimensional Regression
- On the asymptotic variance of the debiased Lasso
- Posterior asymptotic normality for an individual coordinate in high-dimensional linear regression
Cites Work
- Least angle regression. (With discussion)
- Title not available (Why is that?)
- Statistics for high-dimensional data. Methods, theory and applications.
- Simultaneous analysis of Lasso and Dantzig selector
- Confidence Intervals and Hypothesis Testing for High-Dimensional Regression
- Near-Optimal Signal Recovery From Random Projections: Universal Encoding Strategies?
- Regularization and Variable Selection Via the Elastic Net
- Confidence Intervals for Low Dimensional Parameters in High Dimensional Linear Models
- On asymptotically optimal confidence regions and tests for high-dimensional models
- Regularized multivariate regression for identifying master predictors with application to integrative genomics study of breast cancer
- DISTRIBUTION OF EIGENVALUES FOR SOME SETS OF RANDOM MATRICES
- Decoding by Linear Programming
- Compressed sensing
- Title not available (Why is that?)
- Addition of certain non-commuting random variables
- Replica field theory for deterministic models. II. A non-random spin glass with glassy behaviour
- Expectation consistent approximate inference
- Advanced mean field methods. Theory and practice
- Randomly Spread CDMA: Asymptotics Via Statistical Physics
- Cross validation in Lasso and its acceleration
- Analysis of Regularized LS Reconstruction and Random Matrix Ensembles in Compressed Sensing
- Inference for Generalized Linear Models via Alternating Directions and Bethe Free Energy Minimization
Cited In (6)
- Debiasing the Lasso: optimal sample size for Gaussian designs
- Measurement error in Lasso: impact and likelihood bias correction
- The distribution of the Lasso: uniform control over sparse balls and adaptive parameter tuning
- Debiasing the debiased Lasso with bootstrap
- The Lasso with general Gaussian designs with applications to hypothesis testing
- Semi-analytic resampling in Lasso
Uses Software
This page was built for publication: A statistical mechanics approach to de-biasing and uncertainty estimation in Lasso for random measurements
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3303301)
