A Novel Cardiac Attack Prediction and Classification Using Supervised Agent Techniques

In this paper a novel technique is proposed for the prediction of cardiac disease and its classification using intelligent agents. Initially seventy six heart attack symptoms are preprocessed using filter and wrapper based agents .The filter removes the missing or irrelevant symptoms. Wrapper is used to extract the data in the data set according to the expert’s threshold constraints. Dependency of each symptom is identified using dependency checker agent. The classification is based on the prior and posterior probability of the symptoms with the evidence value. As a result classifier agent classify the nature of heart attack patients in to five classes depending on the severity namely absence, starting, mild, moderate, serious. The posterior probability of the class is calculated by the ratio of the product of posterior probability of symptoms and prior probability of the class to the prior probability of the input symptoms. Using the cooperative approach the cardiac problem is solved and verified.