Comparing Nonparametric Versus Parametric Regression Fits

摘要: In general, there will be visible differences between a parametric and nonparametric curve estimate. It is therefore quite natural to compare these in order decide whether the model could justified. An asymptotic quantification distribution of integrated squared difference curves. We show that standard way bootstrapping this statistic fails. use analyse different form for task. call method wild bootstrap apply it fitting Engel curves expenditure data analysis.