Toward a symbiotic expert system for econometric modeling

Abstract

An examination is made of the feasibility of developing an artificially intelligent econometrician as a symbiotic decision support system (SDSS). The authors review the system components of an SDSS and relate them to a DSS for econometric modeling that they have implemented. They present the query language of the PERM (program for econometric modeling) system and offer an extension to the language using process-oriented constructs for model integration. Schemas representing statistical strategies are stored as processes in the extended language and serve as the econometric knowledge base. The authors suggest building an inference processor for this expert system based on experiments to record user query protocols and relate them to the schemas in the knowledge base. Demon constructs in the extended language connect user processes with schemas.<>