28 March 2005 Genetic-program-based data mining for hybrid decision-theoretic algorithms and theories
Author Affiliations +
Abstract
A genetic program (GP) based data mining (DM) procedure has been developed that automatically creates decision theoretic algorithms. A GP is an algorithm that uses the theory of evolution to automatically evolve other computer programs or mathematical expressions. The output of the GP is a computer program or mathematical expression that is optimal in the sense that it maximizes a fitness function. The decision theoretic algorithms created by the DM algorithm are typically designed for making real-time decisions about the behavior of systems. The database that is mined by the DM typically consists of many scenarios characterized by sensor output and labeled by experts as to the status of the scenario. The DM procedure will call a GP as a data mining function. The GP incorporates the database and expert’s rules into its fitness function to evolve an optimal decision theoretic algorithm. A decision theoretic algorithm created through this process will be discussed as well as validation efforts showing the utility of the decision theoretic algorithm created by the DM process. GP based data mining to determine equations related to scientific theories and automatic simplification methods based on computer algebra will also be discussed.
© (2005) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
James F. Smith, James F. Smith, } "Genetic-program-based data mining for hybrid decision-theoretic algorithms and theories", Proc. SPIE 5803, Intelligent Computing: Theory and Applications III, (28 March 2005); doi: 10.1117/12.603151; https://doi.org/10.1117/12.603151
PROCEEDINGS
12 PAGES


SHARE
Back to Top