Data Farming in Support of Military Decision Makers
Abstract
Data farming combines the rapid prototyping capability of certain simulation models with the exploratory power of high performance computing to rapidly generate insight into questions. The aim is to develop a better understanding of landscapes of possibilities as well as outliers that may be discovered through simulation experiments. In this paper we will provide an overview of the overall data farming process as well as discuss methods and techniques that are used within the process. These methods include the application of design of experiments to computational experiments in an iterative process of team-based rapid model prototyping, optimized statistical sampling of the experimental design space, high performance computing, multi-dimensional analysis and visualization, and tools and interfaces for executing these actions. After the concept of data farming was put forth in 1997, the United States Marine Corps utilized the techniques in their Project Albert. This project focused on questions that were fundamental to decision makers, but could not be answered through traditional methods. It relied on the combination of small simulation models, high performance computing, and data farming. During the project, which existed from 1998 to 2006, an international community of interest developed around the topic of data farming. Multi-disciplinary teams of researchers, military officers, and subject matter experts have been using the techniques in collaborative environments since the first international workshop in 1999 and have continued since the project ended.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2009
- Accession Number
- ADA568002
Entities
People
- Daniel Nitsch
- Gary Horne
- Karsten Haymann
- Stephan Seichter
- Susan M. Sanchez
Organizations
- Naval Postgraduate School