Design of experiment

Abstract

In the case of lack of theoretical knowledge of the relationships between input and outputs variables of a system, we have a second possibility: the experimental reconstruction of these relationships. In the case of theoretical knowledge, the system is represented by linear or nonlinear equations between input and output variables, while in the absence of them an experiment must be planned to reach a certain level of knowledge of the system: the Design of experiment (DOE). This tool is general enough to be applied to different types of variables (e.g. categorical variables). Given the introductory character of this chapter, the presentation is limited to basic aspects.