- 期刊
- OpenAccess
modelSampler: An R Tool for Variable Selection and Model Exploration in Linear Regression
並列摘要
We have developed a tool for model space exploration and variable selection in linear regression models based on a simple spike and slab model (Dey, 2012). The model chosen is the best model with minimum final prediction error (FPE) values among all other models. This is implemented via the R package modelSampler.However, model selection based on FPE criteria is dubious and questionable as FPE criteria can be sensitive to perturbations in the data. This R package can be used for empirical assessment of the stability of FPE criteria. A stable model selection is accomplished by using a bootstrap wrapper that calls the primary function of the package several times on the bootstrapped data. The heart of the method is the notion of model averaging for stable variable selection and to study the behavior of variables over the entire model space, a concept invaluable in high dimensional situations.
延伸閱讀
- Wu, J. H. (2015). Improving Model Adaptation for Permutation Problems by RTR with Model-associated Method [master's thesis, National Taiwan University]. Airiti Library. https://doi.org/10.6342/NTU.2015.00303
- Zeng, Y., Wang, X., & Lin, L. (2013). Remodeling and Estimation for Sparse Partially Linear Regression Models. Abstract and Applied Analysis, 2013(), 887-897-887. https://doi.org/10.1155/2013/687151
- Suriya, K., & Niwattisaiwong, S. (2019). Adaptive Logit Model: Enhancement of Predictability in Logistic Regression. International Journal of Intelligent Technologies and Applied Statistics, 12(2), 169-181. https://doi.org/10.6148/IJITAS.201906_12(2).0004
- Dey, T. (2012). A Bimodal Spike and Slab Model for Variable Selection and Model Exploration. Journal of Data Science, 10(3), 363-383. https://doi.org/10.6339/JDS.2012.10(3).1032
- ZHANG, Z., HE, Z., QIN, Y., SHEN, Y., SHIA, B. C., & LI, Y. (2023). Variable Selection with Scalable Bootstrapping in Generalized Linear Model for Massive Data. Journal of Data Science, 21(1), 87-105. https://doi.org/10.6339/22-JDS1052