Prediction of postoperative recovery based on a computational rules extractor

Abstract

One important factor for the patients in a postoperative recovery is hypothermia. The doctor must decide whether the patients should be sent to another place with better medical therapy. We therefore adopt the proposed PSO (particle swarm optimization) based Fuzzy classifier to retrieve the crisp rules from the postoperative given medical data from UCI machine learning database, where the rules can be used to assist in doctor diagnosis. The average correct ratio of our prediction for the postoperative recovery is about 84%.