UNS Conference Portal, The 1st International Conference on Science, Mathematics, Environment and Education 2017

Font Size: 
Comparison of Spline Estimator On Correlation Between Different Levels of Response In Birespon Smoothing Spline Nonparametric Regression
Adji Achmad Rinaldo Fernandes

Last modified: 2017-07-05


Nonparametric regression is one method that can be used to determine the causal relationship between the predictor and the response. The method has the advantage of not requiring the curve shape of known predictors of response forms. Nonparametric regression birespon with smoothing spline is one tool that can be used for research that has two responses were correlated. The purpose of this study is to apply and search for optimal correlation value that can be used on birespon nonparametric smoothing spline regression. The study uses data that has been in the design simulation has nonlinear properties and have two responses. After that it will be done with a nonparametric regression estimation response birespon smoothing spline. Will do a comparison between models to accommodate a correlation and not. The value to be compared is the value of R2 between the two models. The results showed that the optimal correlation obtained is 0.5. This means that the more optimal use nonparametric regression on the correlation between the response is above the value of 0.5.