Arbeitspapier
Double/de-biased machine learning using regularized Riesz representers
We provide adaptive inference methods for linear functionals of L1-regularized linear approximations to the conditional expectation function. Examples of such functionals include average derivatives, policy effects, average treatment effects, and many others. The construction relies on building Neyman-orthogonal equations that are approximately invariant to perturbations of the nuisance parameters, including the Riesz representer for the linear functionals. We use L1-regularized methods to learn the approximations to the regression function and the Riesz representer, and construct the estimator for the linear functionals as the solution to the orthogonal estimating equations. We establish that under weak assumptions the estimator concentrates in a 1/vn neighborhood of the target with deviations controlled by the normal laws, and the estimator attains the semi-parametric efficiency bound in many cases. In particular, either the approximation to the regression function or the approximation to the Rietz representer can be "dense" as long as one of them is sufficiently "sparse". Our main results are non-asymptotic and imply asymptotic uniform validity over large classes of models.
- Language
-
Englisch
- Bibliographic citation
-
Series: cemmap working paper ; No. CWP15/18
- Classification
-
Wirtschaft
- Subject
-
Approximate Sparsity vs. Density
Double/De-biased Machine Learning
Regularized Riesz Representers
Linear Functionals
- Event
-
Geistige Schöpfung
- (who)
-
Chernozhukov, Victor
Newey, Whitney K.
Robins, James
- Event
-
Veröffentlichung
- (who)
-
Centre for Microdata Methods and Practice (cemmap)
- (where)
-
London
- (when)
-
2018
- DOI
-
doi:10.1920/wp.cem.2018.1518
- Handle
- Last update
-
10.03.2025, 11:43 AM CET
Data provider
ZBW - Deutsche Zentralbibliothek für Wirtschaftswissenschaften - Leibniz-Informationszentrum Wirtschaft. If you have any questions about the object, please contact the data provider.
Object type
- Arbeitspapier
Associated
- Chernozhukov, Victor
- Newey, Whitney K.
- Robins, James
- Centre for Microdata Methods and Practice (cemmap)
Time of origin
- 2018
Other Objects (12)
Adversarial estimation of Riesz representers
High-dimensional econometrics and regularized GMM
Plug-in regularized estimation of high-dimensional parameters in nonlinear semiparametric models
Simultaneous inference for best linear predictor of the conditional average treatment effect and other structural functions
Posterior inference in curved exponential families under increasing dimensions
On the computational complexity of MCMC-based estimators in large samples
Post-l1-penalized estimators in high-dimensional linear regression models
Quantile models with endogeneity
Inference for extremal conditional quantile models, with an application to market and birthweight risks
L1-Penalised quantile regression in high-dimensional sparse models
Posterior inference in curved exponential families under increasing dimensions
Double machine learning for treatment and causal parameters
Adversarial estimation of Riesz representers
High-dimensional econometrics and regularized GMM
Plug-in regularized estimation of high-dimensional parameters in nonlinear semiparametric models
Simultaneous inference for best linear predictor of the conditional average treatment effect and other structural functions
Posterior inference in curved exponential families under increasing dimensions
On the computational complexity of MCMC-based estimators in large samples
Post-l1-penalized estimators in high-dimensional linear regression models
Quantile models with endogeneity
Inference for extremal conditional quantile models, with an application to market and birthweight risks
L1-Penalised quantile regression in high-dimensional sparse models
Posterior inference in curved exponential families under increasing dimensions
Double machine learning for treatment and causal parameters
Adversarial estimation of Riesz representers
High-dimensional econometrics and regularized GMM
Plug-in regularized estimation of high-dimensional parameters in nonlinear semiparametric models
Simultaneous inference for best linear predictor of the conditional average treatment effect and other structural functions
Posterior inference in curved exponential families under increasing dimensions
On the computational complexity of MCMC-based estimators in large samples
Post-l1-penalized estimators in high-dimensional linear regression models
Quantile models with endogeneity
Inference for extremal conditional quantile models, with an application to market and birthweight risks
L1-Penalised quantile regression in high-dimensional sparse models
Posterior inference in curved exponential families under increasing dimensions