Variable importance based interaction modelling with an application on initial spread of COVID-19 in China-Reference-Cited by-同舟云学术

Variable importance based interaction modelling with an application on initial spread of COVID-19 in China

Published:2024-07-11 Issue: Volume: Page:
ISSN:0035-9254
Container-title:Journal of the Royal Statistical Society Series C: Applied Statistics
language:en
Short-container-title:

Author:

Zhang Jianqiang¹²,Chen Ze³,Yang Yuhong⁴,Xu Wangli²^ORCID

Affiliation:

1. School of Statistics and Mathematics, Zhejiang Gongshang University , Hangzhou , China

2. Center for Applied Statistics and School of Statistics, Renmin University of China , Beijing , China

3. Zhongtai Securities Institute for Financial Studies, Shandong University , Jinan , China

4. Yau Mathematical Sciences Center, Tsinghua University , Beijing , China

Abstract

Abstract Interaction selection for linear regression is useful in many fields of modern science, yet very challenging. Existing methods focus on finding one optimal model but they may perform poorly in terms of stability for high-dimensional data, and they do not typically deal with categorical predictors. In this paper, we introduce a variable importance based interaction modelling (VIBIM) procedure for learning interactions in a linear regression model with both continuous and categorical predictors. We apply the VIBIM procedure to a COVID-19 data and show that the VIBIM approach leads to better models in terms of interpretability, stability, reliability, and prediction.

Funder

Beijing Natural Science Foundation

National Natural Science Foundation of China

Renmin University of China

MOE Project of Key Research Institute of Humanities and Social Sciences

Postdoctoral Fellowship Program of CPSF

Publisher

Oxford University Press (OUP)

Link

https://academic.oup.com/jrsssc/advance-article-pdf/doi/10.1093/jrsssc/qlae029/58504144/qlae029.pdf

Reference55 articles.

1. Risk bounds for model selection via penalization;Barron;Probability Theory and Related Fields,1999

2. Conditional sure independence screening;Barut;Journal of the American Statistical Association,2016

3. Coordinate descent algorithms for nonconvex penalized regression, with applications to biological feature selection;Breheny;Annals of Applied Statistics,2011

4. Heuristics of instability and stabilization in model selection;Breiman;The Annals of Statistics,1996

5. Random forests;Breiman;Machine Learning,2001