Predictive inference with random forests: A new perspective on classical analyses

Predictive inference with random forests: A new perspective on classical analyses

Despite the number of problems that can occur when core model assumptions are violated, nearly all quantitative political science research relies on inflexible regression models that require a linear relationship between dependent and independent variables for valid inference. We argue that nonparam...

Full description

Bibliographic Details
Main Authors: Richard J. McAlexander, Lucas Mentch
Format: Article
Language:English
Published: SAGE Publishing 2020-02-01
Series:Research & Politics
Online Access:https://doi.org/10.1177/2053168020905487
id doaj-9c338c60ff2d450292c9701be83a5f20
record_format Article
spelling doaj-9c338c60ff2d450292c9701be83a5f202020-11-25T03:56:12ZengSAGE PublishingResearch & Politics2053-16802020-02-01710.1177/2053168020905487Predictive inference with random forests: A new perspective on classical analysesRichard J. McAlexander0Lucas Mentch1Columbia University, USAUniversity of Pittsburgh, USADespite the number of problems that can occur when core model assumptions are violated, nearly all quantitative political science research relies on inflexible regression models that require a linear relationship between dependent and independent variables for valid inference. We argue that nonparametric statistical learning methods like random forests are capable of combining the benefits of interpretability and flexibility. Recent work has shown that under suitable regularity conditions, averaging over predictions made by subsampled random forests produces asymptotically normal predictions. After estimating the variance, this property can be exploited to produce hypothesis tests and confidence intervals analogous to those produced within a parametric framework. We demonstrated the utility of this approach by replicating an important study on the determinants of civil war onset and show that subtle nonlinear relationships are uncovered, providing a new perspective on these ongoing research questions.https://doi.org/10.1177/2053168020905487
collection DOAJ
language English
format Article
sources DOAJ
author Richard J. McAlexander
Lucas Mentch
spellingShingle Richard J. McAlexander
Lucas Mentch
Predictive inference with random forests: A new perspective on classical analyses
Research & Politics
author_facet Richard J. McAlexander
Lucas Mentch
author_sort Richard J. McAlexander
title Predictive inference with random forests: A new perspective on classical analyses
title_short Predictive inference with random forests: A new perspective on classical analyses
title_full Predictive inference with random forests: A new perspective on classical analyses
title_fullStr Predictive inference with random forests: A new perspective on classical analyses
title_full_unstemmed Predictive inference with random forests: A new perspective on classical analyses
title_sort predictive inference with random forests: a new perspective on classical analyses
publisher SAGE Publishing
series Research & Politics
issn 2053-1680
publishDate 2020-02-01
description Despite the number of problems that can occur when core model assumptions are violated, nearly all quantitative political science research relies on inflexible regression models that require a linear relationship between dependent and independent variables for valid inference. We argue that nonparametric statistical learning methods like random forests are capable of combining the benefits of interpretability and flexibility. Recent work has shown that under suitable regularity conditions, averaging over predictions made by subsampled random forests produces asymptotically normal predictions. After estimating the variance, this property can be exploited to produce hypothesis tests and confidence intervals analogous to those produced within a parametric framework. We demonstrated the utility of this approach by replicating an important study on the determinants of civil war onset and show that subtle nonlinear relationships are uncovered, providing a new perspective on these ongoing research questions.
url https://doi.org/10.1177/2053168020905487
work_keys_str_mv AT richardjmcalexander predictiveinferencewithrandomforestsanewperspectiveonclassicalanalyses
AT lucasmentch predictiveinferencewithrandomforestsanewperspectiveonclassicalanalyses
_version_ 1724466549165654016

Similar Items