Bias and variance residuals for machine learning nonlinear simplex regressions
作者：
Highlights：
• Nonlinear simplex regression models can be used for supervised machine learning.
• We propose two new residuals for a supervised nonlinear simplex regressions.
• The new residuals were obtained from the joint estimation of all model parameters.
• The new residuals do not require computation of large projection matrices.
• Three empirical applications are presented; they favor the new residuals.
摘要
•Nonlinear simplex regression models can be used for supervised machine learning.•We propose two new residuals for a supervised nonlinear simplex regressions.•The new residuals were obtained from the joint estimation of all model parameters.•The new residuals do not require computation of large projection matrices.•Three empirical applications are presented; they favor the new residuals.
论文关键词：Double bounded variable,Machine learning,Outlier,Simplex regression
论文评审过程：Received 14 May 2020, Revised 4 March 2021, Accepted 22 July 2021, Available online 30 July 2021, Version of Record 4 August 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.115656
