Fuzzy beans in classification

Research highlights? New fuzzy bean based classification method is given. ? For optimization of suitable fuzzy bean and ideal vectors Differential Evolution algorithm was utilized. ? Promising results in mean classification accuracy was gained with proposed method and used medical data sets compared to previous results. In the article fuzzy bean based classifier is given. Three different type of structures are tested in the classifier. Fuzzy bean based classifier is supervised learning method and with proper optimization scheme promising results were covered. Differential evolution algorithm was used in optimizing the required parameters. Classifier was applied to diagnosis of liver-disorders with 73.9% classification accuracy, to Pima-Indian diabetes with 77.8% accuracy, to diagnosis of breast cancer with 97.8% accuracy and to echocardiogram with 93.0% accuracy. Results can be considered good and they compare well with results reported with literature.