Choosing among regularized estimators in empirical economics: the risk of machine learning
Many settings in empirical economics involve estimation of a large number of parameters. In such settings, methods that combine regularized estimation and data-driven choices of regularization parameters are useful. We provide guidance to applied researchers on the choice between regularized estimat...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
MIT Press - Journals
2020
|
| Online Access: | https://hdl.handle.net/1721.1/124406 |
| _version_ | 1826196910362329088 |
|---|---|
| author | Abadie, Alberto Kasy, Maximilian |
| author2 | Massachusetts Institute of Technology. Department of Economics |
| author_facet | Massachusetts Institute of Technology. Department of Economics Abadie, Alberto Kasy, Maximilian |
| author_sort | Abadie, Alberto |
| collection | MIT |
| description | Many settings in empirical economics involve estimation of a large number of parameters. In such settings, methods that combine regularized estimation and data-driven choices of regularization parameters are useful. We provide guidance to applied researchers on the choice between regularized estimators and data-driven selection of regularization parameters. We characterize the risk and relative performance of regularized estimators as a function of the data-generating process and show that data-driven choices of regularization parameters yield estimators with risk uniformly close to the risk attained under the optimal (unfeasible) choice of regularization parameters. We illustrate using examples from empirical economics.©2019 |
| first_indexed | 2024-09-23T10:39:34Z |
| format | Article |
| id | mit-1721.1/124406 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T10:39:34Z |
| publishDate | 2020 |
| publisher | MIT Press - Journals |
| record_format | dspace |
| spelling | mit-1721.1/1244062022-09-30T22:06:58Z Choosing among regularized estimators in empirical economics: the risk of machine learning Abadie, Alberto Kasy, Maximilian Massachusetts Institute of Technology. Department of Economics Many settings in empirical economics involve estimation of a large number of parameters. In such settings, methods that combine regularized estimation and data-driven choices of regularization parameters are useful. We provide guidance to applied researchers on the choice between regularized estimators and data-driven selection of regularization parameters. We characterize the risk and relative performance of regularized estimators as a function of the data-generating process and show that data-driven choices of regularization parameters yield estimators with risk uniformly close to the risk attained under the optimal (unfeasible) choice of regularization parameters. We illustrate using examples from empirical economics.©2019 2020-03-30T14:23:11Z 2020-03-30T14:23:11Z 2019-12 2017-12 2019-10-22T14:45:25Z Article http://purl.org/eprint/type/JournalArticle 1530-9142 0034-6535 https://hdl.handle.net/1721.1/124406 Abadie, Alberto and Maximilian Kasy, "Choosing among regularized estimators in empirical economics: the risk of machine learning." Review of economics and statistics 101, 5 (December 2019): p. 743-62 doi 10.1162/rest_a_00812 ©2019 Authors en 10.1162/REST_A_00812 Review of economics and statistics Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf MIT Press - Journals MIT Press |
| spellingShingle | Abadie, Alberto Kasy, Maximilian Choosing among regularized estimators in empirical economics: the risk of machine learning |
| title | Choosing among regularized estimators in empirical economics: the risk of machine learning |
| title_full | Choosing among regularized estimators in empirical economics: the risk of machine learning |
| title_fullStr | Choosing among regularized estimators in empirical economics: the risk of machine learning |
| title_full_unstemmed | Choosing among regularized estimators in empirical economics: the risk of machine learning |
| title_short | Choosing among regularized estimators in empirical economics: the risk of machine learning |
| title_sort | choosing among regularized estimators in empirical economics the risk of machine learning |
| url | https://hdl.handle.net/1721.1/124406 |
| work_keys_str_mv | AT abadiealberto choosingamongregularizedestimatorsinempiricaleconomicstheriskofmachinelearning AT kasymaximilian choosingamongregularizedestimatorsinempiricaleconomicstheriskofmachinelearning |