Damien Perret, R&D Scientist, CEA
François Bergeret, Senior Statistician, Ippon Innovation
Carole Soual, Chief Statistician, Ippon Innovation
Isabelle Giboire, Isabelle, CEA
JMP software was initially implemented at CEA in 2010 by R&D teams who develop nuclear glass formulation. Over the years, JMP has been used for multiple purposes, such as data visualization of highly complex composition domains, optimal mixture designs, and machine learning techniques to create property-to-composition predictive models. More recently, JMP enabled us to develop very innovative methodologies. Two case studies will be presented. First, we will show an original approach based on an automatic and intelligent subsampling of the data, combining techniques of optimal designs and several predictive methods in JMP and JMP Pro to create very robust and accurate predictive models. Second, we will present an amazing benefit of using the Simulation platform where a response is below the limit of detection in the most part of the design space.
