Geometrically designed variable knot splines in generalized (non-)linear models
作者：
Highlights：
• We propose an extension/generalization of the GeDS methodology, recently developed by Kaishev et al. (2016) for the Normal univariate spline regression case.
• To the more general GNM/GLM context.
• To the multivariate case of more than one independent variable.
• An extensive comparison of GeDS with major existing spline fitting methodologies is presented.
• The proposed generalized GeDS methodology is implemented in an R package, named GeDS, which is available from the Comprehensive R Archive Network (CRAN) at http://CRAN.R-project.org/package=GeDS.
摘要
•We propose an extension/generalization of the GeDS methodology, recently developed by Kaishev et al. (2016) for the Normal univariate spline regression case.•To the more general GNM/GLM context.•To the multivariate case of more than one independent variable.•An extensive comparison of GeDS with major existing spline fitting methodologies is presented.•The proposed generalized GeDS methodology is implemented in an R package, named GeDS, which is available from the Comprehensive R Archive Network (CRAN) at http://CRAN.R-project.org/package=GeDS.
论文关键词：Variable-knot spline regression,Tensor product B-splines,Greville abscissae,Control polygon,Generalized non-linear models
论文评审过程：Received 17 November 2021, Revised 6 August 2022, Accepted 17 August 2022, Available online 7 September 2022, Version of Record 7 September 2022.
论文官网地址：https://doi.org/10.1016/j.amc.2022.127493
