Genetic Programming Approach to Data Mining and Knowledge Discovery from Store Separation Trajectories

Development of a data mining and knowledge discovery software package for store separation trajectory analysis is proposed. Multiple tools will be included in the software to allow a spectrum of investigations such as visual data mining, uncertainty analysis, worst-case analysis and variation analysis. These tools can be used by store separation trajectory analysts to gain a deeper understanding of the information contained in the trajectories from sources such as Monte-Carlo simulations and Genetic Algorithm searches. The software will help extract important trends in the data and direct the analyst's attention to areas requiring further analysis. Phase I feasibility demonstration project developed a preliminary prototype of the software that works in a well-known numerical environment with an open-source database managing the store separation data. The software features were demonstrated with respect to the store separation analysis workflow. The Phase II research will refine the analytical tools to more specifically focus on the issues of interest to the Air Force. A commercial quality software package capable of running under multiple operating systems will be developed and delivered to the Air Force with all the pertinent documentation at the end of the proposed research.