Forecasting System for Heart Diseases: A Classification Approach

The medical field provides enormous quantities of data that contain an unobserved pattern that can be useful for decisions. This paper intended to develop a forecasting system to detect the early existence of heart diseases based on fourteen variables from patients' historical data. The system used the naïve bayes data mining technique to analyze the inputted medical data. This paper used the Rapid Application Development (RAD) Software Life Cycle in designing and developing the system. The system was simulated using synthetic datasets namely, Cleveland and Statlog. Results of this study showed that the system provided adequate features, the dataset input training page, data verification page, and forecasting results page for predicting heart diseases.