A MACHINE-LEARNING APPROACH TO AUTOMATED KNOWLEDGE-BASE BUILDING FOR REMOTE-SENSING IMAGE-ANALYSIS WITH GIS DATA

A machine learning approach to automated building of knowledge bases f or image analysis expert systems incorporating GIS data is presented. The method uses an inductive learning algorithm to generate production rules from training data. With this method, building a knowledge base for a rule-based expert system is easier than using the conventional knowledge acquisition approach. The knowledge base built by this metho d was used by an expert system to perform a wetland classification of Par Pond on the Savannah River Site, South Carolina using SPOT multisp ectral imagery and GIS data. To evaluate the performance of the result ant knowledge base, the classification result was compared to classifi cations with two conventional methods. The accuracy assessment and the analysis of the resultant production rules suggest that the knowledge base built by the machine learning method was of good quality for ima ge analysis with GIS data.