The Bootstrap and Beyond: Using JSL to Implement Additional Resampling Techniques
Michael Crotty, PhD, JMP Research Statistician, SAS
Clay Barker, PhD, JMP Research Statistician, SAS
Bootstrap is a powerful resampling technique for measuring the accuracy of statistical estimators and for making decisions. Resampling techniques are especially useful when we do not know the distribution of our statistic or when we do not trust the underlying assumptions of our model. JMP 10 introduced a "one-click bootstrap" feature that allows one to use the nonparametric bootstrap to help make decisions in a variety of settings. But why stop there? With a little bit of scripting, we can implement more resampling techniques like the parametric bootstrap for decision making and bootstrap aggregating (bagging) for prediction. After a brief introduction to the nonparametric bootstrap and its use in JMP, we will introduce additional resampling techniques and discuss how to implement them using JSL.
