On efficiency properties of an R-square coefficient based on final prediction error
暂无分享,去创建一个
Abstract This paper studies the efficiency of the coefficient of determination ( R F P E 2 ) based on final prediction error which is proposed by Rousson and Gosoniu (2007) , and compares it with conventional goodness-of-fit measures ( R 2 , R a 2 ) in linear regression models with both normal and non-normal disturbances.
[1] Valentin Rousson,et al. An R-square coefficient based on final prediction error , 2007 .
[2] Aman Ullah,et al. Finite Sample Econometrics , 2004 .
[3] Anil K. Srivastava,et al. The coefficient of determination and its adjusted version in linear regression models , 1995 .
[4] J. Cramer,et al. Mean and variance of R2 in small and moderate samples , 1987 .