Oracle inequalities for sign constrained generalized linear models

Yuta Koike, Yuta Tanoue

Research output: Contribution to journalArticle

Abstract

High-dimensional data have recently been analyzed because of advancements in data collection technology. Although many methods have been developed for sparse recovery in the past 20 years, most of these methods require the selection of tuning parameters. This requirement means that the results obtained with these methods heavily depend on tuning. Theoretical properties are developed for sign-constrained generalized linear models with convex loss function, which is one of the sparse regression methods that does not require tuning parameters. Recent studies on this subject have shown that, in the case of linear regression, sign-constraints alone could be as efficient as the oracle method if the design matrix enjoys a suitable assumption in addition to a traditional compatibility condition. This type of result is generalized to a model that encompasses logistic and quantile regressions. Some numerical experiments are performed to confirm the theoretical findings.

Original languageEnglish
JournalEconometrics and Statistics
DOIs
Publication statusPublished - 2019 Jan 1
Externally publishedYes

Fingerprint

Oracle Inequalities
Generalized Linear Model
Parameter Tuning
Quantile Regression
Compatibility Conditions
High-dimensional Data
Loss Function
Logistic Regression
Linear regression
Convex function
Tuning
Recovery
Regression
Numerical Experiment
Generalized linear model
Requirements

Keywords

  • High-dimensions
  • Oracle inequality
  • Sign-constraints
  • Sparsity

ASJC Scopus subject areas

  • Statistics and Probability
  • Economics and Econometrics
  • Statistics, Probability and Uncertainty

Cite this

Oracle inequalities for sign constrained generalized linear models. / Koike, Yuta; Tanoue, Yuta.

In: Econometrics and Statistics, 01.01.2019.

Research output: Contribution to journalArticle

@article{4b68749477a84eaea77ba919c428b123,
title = "Oracle inequalities for sign constrained generalized linear models",
abstract = "High-dimensional data have recently been analyzed because of advancements in data collection technology. Although many methods have been developed for sparse recovery in the past 20 years, most of these methods require the selection of tuning parameters. This requirement means that the results obtained with these methods heavily depend on tuning. Theoretical properties are developed for sign-constrained generalized linear models with convex loss function, which is one of the sparse regression methods that does not require tuning parameters. Recent studies on this subject have shown that, in the case of linear regression, sign-constraints alone could be as efficient as the oracle method if the design matrix enjoys a suitable assumption in addition to a traditional compatibility condition. This type of result is generalized to a model that encompasses logistic and quantile regressions. Some numerical experiments are performed to confirm the theoretical findings.",
keywords = "High-dimensions, Oracle inequality, Sign-constraints, Sparsity",
author = "Yuta Koike and Yuta Tanoue",
year = "2019",
month = "1",
day = "1",
doi = "10.1016/j.ecosta.2019.02.001",
language = "English",
journal = "Econometrics and Statistics",
issn = "2452-3062",
publisher = "Elsevier BV",

}

TY - JOUR

T1 - Oracle inequalities for sign constrained generalized linear models

AU - Koike, Yuta

AU - Tanoue, Yuta

PY - 2019/1/1

Y1 - 2019/1/1

N2 - High-dimensional data have recently been analyzed because of advancements in data collection technology. Although many methods have been developed for sparse recovery in the past 20 years, most of these methods require the selection of tuning parameters. This requirement means that the results obtained with these methods heavily depend on tuning. Theoretical properties are developed for sign-constrained generalized linear models with convex loss function, which is one of the sparse regression methods that does not require tuning parameters. Recent studies on this subject have shown that, in the case of linear regression, sign-constraints alone could be as efficient as the oracle method if the design matrix enjoys a suitable assumption in addition to a traditional compatibility condition. This type of result is generalized to a model that encompasses logistic and quantile regressions. Some numerical experiments are performed to confirm the theoretical findings.

AB - High-dimensional data have recently been analyzed because of advancements in data collection technology. Although many methods have been developed for sparse recovery in the past 20 years, most of these methods require the selection of tuning parameters. This requirement means that the results obtained with these methods heavily depend on tuning. Theoretical properties are developed for sign-constrained generalized linear models with convex loss function, which is one of the sparse regression methods that does not require tuning parameters. Recent studies on this subject have shown that, in the case of linear regression, sign-constraints alone could be as efficient as the oracle method if the design matrix enjoys a suitable assumption in addition to a traditional compatibility condition. This type of result is generalized to a model that encompasses logistic and quantile regressions. Some numerical experiments are performed to confirm the theoretical findings.

KW - High-dimensions

KW - Oracle inequality

KW - Sign-constraints

KW - Sparsity

UR - http://www.scopus.com/inward/record.url?scp=85062183255&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85062183255&partnerID=8YFLogxK

U2 - 10.1016/j.ecosta.2019.02.001

DO - 10.1016/j.ecosta.2019.02.001

M3 - Article

AN - SCOPUS:85062183255

JO - Econometrics and Statistics

JF - Econometrics and Statistics

SN - 2452-3062

ER -