A Full Explanation Facility for an MLP Network That Classifies Low-Back-Pain Patients and for Predicting MLP Reliability

This chapter presents a full explanation facility for any standard MLP network with binary input neurons that performs a classification task. The interpretation of any input case is represented by a non-linear ranked data relationship of key inputs. The knowledge that the MLP has learned is represented by ranked class profiles or as a set of rules. The explanation facility discovers the MLP knowledge bounds, enabling novelty detection to be implemented and the predictability of the MLP to be investigated. Results using the facility are presented for a 48-dimensional real world MLP that classifies low-back-pain patients.