The Use of Observational Data to Implement an Optimal Experimental Design

This paper is focused on observational data and on the use of large data sets to implement an experimental design without additional runs, for an efficient use of these data. More specifically, the proposed procedure is based on several steps, aimed at avoiding some problems, such as lack of randomization and efficiency, obtaining a final experimental design that can be assumed as optimal from the point of view of a sequential optimality.