INTERPRETATION AND KNOWLEDGE DISCOVERY FROM THE MULTILAYER PERCEPTRONNETWORK - OPENING THE BLACK-BOX

This paper interprets the outputs from the multilayer perceptron (MLP) network by finding the input data features at the input layer of the network which activate the hidden layer feature detectors. This leads directly to the deduction of the significant data inputs, the inputs t hat the network actually uses to perform the input/output mapping for a classification task, and the discovery of the most significant of th ese data inputs. The analysis presents a method for providing explanat ions for the network outputs and for representing the knowledge learne d by the network in the form of significant input data relationships. During network development the explanation facilities and data relatio nships can be used for network validation and verification, and after development, for rule induction and data mining where this method prov ides a potential tool for knowledge discovery in databases (KDD).