Data Farming: The Meanings and Methods Behind the Metaphor

Abstract

Data farming captures the notion of purposeful data generation from simulation models. The ready availability of computing power has fundamentally changed the way simulation and other computational models can be used to provide insights to decision makers. Large-scale designed experiments let us grow the simulation output efficiently and effectively. We can explore massive input spaces, use statistical and visualization techniques to uncover interesting features of complex response surfaces, and explicitly identify cause-and-effect relationships. Nonetheless, there are many opportunities for research methods that could further enhance this process. I will begin with a brief overview of key differences between physical and simulation experiments, as well as current data farming capabilities and their relationship to emerging techniques in data science and analytics. I will then share some thoughts about opportunities and challenges for further improving the state of the art, and transforming the state of the practice, in this domain.