Design of experiments (DOE) is a data-driven strategy that helps practitioners to better understand the underlying behavior of a system by identifying those factors that most affect desired outcomes. By using optimal DOE, you can more efficiently take account of your prior knowledge and commit resources that are commensurate with what you want to learn.
Join author and University of Leuven Professor Peter Goos and SAS Principal Research Fellow Bradley Jones to learn how to use optimal designs that can reduce the cost of experimentation, accommodate multiple types of factors, and lead to optimal factor settings even when the design space is constrained.
The seminar is based on real industrial case studies from Optimal Design of Experiments, by Jones and Goos, and addresses the use of modern, computer-based methods to:
- Find the few factors that most affect the response of interest.
- Resolve ambiguity about what model best describes the underlying behavior of the system.
- Deal with a problem wherein the blocks cannot be orthogonal to the other factor effects in the model.
- Investigate the behavior of a chemical reaction using a full cubic model in two factors -- even when many of the factor-level combinations are known in advance to be infeasible.
- Take covariate information into account when setting up an experiment.