Publication
IBM Systems Journal
Paper
A probabilistic estimation framework for predictive modeling analytics
Abstract
A probabilistic estimation framework for predictive modeling analytics was presented. IBM ProbE is an extensible, embeddable and scalable modeling engine for implementing segmentation-based modeling techniques. Here the data records are partitioned into segments and separate predictive models are developed for each segments. It ignores the strong influence that segmentation exerts on the predictive accuracies of the models within each segment.