The analysis of large data sets, where hundreds to thousands of measurements on a part, process, or sample are taken requires innovative approaches. The Predictor Screening platform provides a method of screening many predictors for their ability to predict an outcome. For example, predictor screening can be used to help identify biomarkers from thousands tested in samples from patients with and without a condition to predict the condition.

Predictor screening differs from response screening. Response screening tests factors one at a time as a predictor of the response. Predictor screening uses bootstrap forest partitioning to evaluate the contribution of predictors on the response. The partition models are built on multiple predictors. Predictor screening can identify predictors that might be weak alone but strong when used in combination with other predictors. See “Response Screening” chapter for details.