A Connectionist Approach to Producing Rules Describing Monthly UK Divisia Data

This paper demonstrates a mechanism whereby rules can be extracted from a feedforward neural network trained to characterize the money­ price relationship, defined as the relationship between the rate of growth of the money supply and inflation. Monthly Divisia component data is encoded and used to train a group of candidate connectionist architectures. One candidate is selected for rule extraction, using a custom decompositional extraction algorithm that generates rules in human-readable and machine-executable form. Rule and network accuracy are compared, and comments are made on the relationships expressed within the discov­ ered rules. The types of discovered relationships could be used to guide monetary policy decisions.