Learning Engineering Models with the Minimum Description Length Principle

R. Bharat Rao, Stephen C-Y. Lu

This paper discusses discovery of mathematical models from engineering data sets. KEDS, a Knowledge-based Equation Discovery System, identifies several potentially overlapping regions in the problem space, each associated with an equation of different complexity and accuracy. The minimum description length principle, together with the KEDS algorithm, is used to guide the partitioning of the problem space. The KEDS-MDL algorithm has been tested on discovering models for predicting the performance efficiencies of an internal combustion engine.

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.