This paper describes an expert system devoted to long term interest rate forecasting. The rule base is not derived from any econometric model but from the reasoning as made by financial managers. The knowledge base is modelled using an object oriented approach. Based on historical data, certainty factors are used to allow the coexistence of rules which could seem to be contradictory. Beyond a decidability threshold, the proposed rules constitute a good forecasting model.
AKOKA, J. et COMYN-WATTIAU, I. (1997). An Expert System for Long Term Interest Rate Forecasting. Dans: EXPERSYS-97 Artificial Intelligence Applications. IITT, pp. 117-124.