Paper Title

Abstract - This research seeks to use knowledge acquired in hidden dataset to train data mining algorithms for the prediction of heart disease, ensuring effective prediction and minimizing the stress and cost on medical practitioners and patients. The research hope to develop a Cardio-vascular disease prediction system that will be guided using four data mining classification techniques to mine available databases to unravel hidden patterns that can be useful in the effort to understand and develop medical solutions to cardio-vascular disease. The KDD process was the method utilized in carrying out this research. The data used was obtained from a publicly available dataset in the database of the UCI (University of California, Irvine C.A) machine learning repository. Cleveland Clinic Foundation dataset containing 303 patients record and Hungarian Institute of Cardiology dataset containing 294 patient records were integrated selecting 13 relevant attributes from 74. Four data mining algorithms namely Naïve Bayes, Decision Tree, K-Nearest Neighbor, and the Artificial Neural Network were used to mine patterns and knowledge from the dataset. The results obtained shows that the algorithms performed averagely well with a high accuracy of 91.8% for Artificial Neural Network (ANN) and the lowest accuracy of 68.85% for K-Nearest Neighbor. It is therefore recommended that the ANN is the most suitable data mining algorithm to be applied to identifying new cases of patients with cardio-vascular diseases Keywords - Cardio-Vascular Diseases Mining, Classification Algorithms in Healthcare, Artificial Neural Network, Naïve Bayes, Decision Tree, K-Nearest Neighbor. Component