Spline Regression Modelling for Health Problem: Study of Knot and Confidence Interval

Abstract

This article aimed to study about knot and confidence interval for health science using spline nonparameteric regression.The study used simulation and real data. The result showed that numbers of knot should be placed according to the quantil variable in order to get a good estimation of the data function. Inadditon,confidence interval using bayesian and bootstrap method gave no different result for a small sample size whereas for a big sample size bootstrap gave narrower interval.