A New Method for Fish disease Diagnosis System Based on Rough Set and Classifier Fusion

A model of fish disease diagnosis was proposed by combining rough sets theory (RST) with classifier fusion. On the basis of the attribute reduction of RST, the remaining condition attributes were used for the inputs of individual classifiers and the decision attributes as the outputs. The application of Ordered Weighted Averaging (OWA) operator as a classifier fusion approach has been adopted to combine the decisions of four underlying individual classifiers. By using data gathered from reduction fish disease diagnosis case database, the accuracy of OWA-based classifier fusion system has been compared with the individual classifiers. The experiment results show that the model is effective and practicality.