Medical Data Analysis Based on Transparent Classifier

Abstract

This paper investigates the application of transparent classifier to medical data retrieved from the UCI ML repository. The advantage of transparent classifier compared to those based on neural networks is the ability of explaining the result of classification. The process by transparent classifier is to obtain the patterns first through data association in the training data, use the identified patterns to classify the test data by calculating the positive and the negative score, and then determine the class according to the scores.