A Ridge Analysis with Noise Variables

• Activity Number: 184
• Type: Contributed
• Date/Time: Tuesday, August 13, 2002 : 8:30 AM to 10:20 AM
• Sponsor: Section on Quality & Productivity*
• Author(s): John Peterson*+ and Andrew Kuhn
• Affiliation(s): GlaxoSmithKline, Inc. and Becton Dickinson Microbiology Systems
• Keywords: Mixture Experiment ; Response Surface Methodology

Abstract

Ridge analysis is a graphical and inferential method for exploring optimum factor levels of a response surface at fixed distances from the center of the experimental design. This paper proposes an approach to doing a ridge analysis for optimizing a response surface in the presence of noise variables. We extend the approach of Peterson (1993) to include some of the factors as noise variables. This approach allows the investigator to explore factor combinations that lower the mean square error about a target value, while at the same time keeping track of how much the mean response differs from the target value. We also propose a modification of our approach which can be used for "larger is better" or "smaller is better" experiments. We illustrate the proposed method using two examples, one of which is a mixture experiment.