Clay Barker, Principal Research Statistician Developer at JMP
Date: Nov. 9th, 3pm – 4pm CET
"Less is more" may seem like a trite expression, but it is a useful motto when it comes to building regression models. As data sets get bigger and we may have hundreds of predictors or more, powerful tools to help us choose a subset of predictors for explaining our response variable are more important than ever before. In this webinar, we will introduce a handful of automated variable selection techniques like forward selection and the Lasso. Attendees will also learn how to evaluate regression models using information criteria (like the AIC) and cross validation. The Generalized Regression platform in JMP Pro will be used to demonstrate the techniques on example data sets.
Clay Barker is a Principal Research Statistician Developer for JMP, a subsidiary of SAS. While his focus has been on developing variable selection tools for generalized linear models, Clay has also developed tools in for a variety of analyses including nonlinear regression, wavelets, and clustering. Clay earned his PhD in statistics from North Carolina State University.