Detecting major disease in public hospital using ensemble techniques

Hepatitis is chronic disease that becomes major problem in developing countries. Health experts estimate that more than 185 billion people have chronic hepatitis worldwide. This paper attempts to detect major disease such as hepatitis in public hospital using ensemble methods. Several ensemble techniques were applied to acquire knowledge from patient medical records. Afterwards, rule extraction from decision tree and neural network are summarized in order to assist experts in detecting hepatitis. Accuracy of those algorithms is also performed and from the experimental result shows that Bagging, with decision tree as base-classifier, denotes best performance among other classifiers.