부산대학교 도서관

Health-care organizations collect huge amounts of data, which might include certain hidden facts that could be utilized to make better decisions. Some sophisticated data mining methods are utilized to provide suitable findings and make good data judgements. In this study, affinity propagation clustering & neural networks are utilized to build an adequate heart disease predicting system for anticipating the risk level of heart disease. The algorithm predicts using 15 medical criteria such as age, gender, blood pressure, diabetes, and obesity. The EHDPS predicts a patient's risk of developing heart disease. It enables the generation of a large amount of data, such as associations between medical indicators related to heart disease as well as patterns. A multilayer neural network containing back propagation has been employed as the training approach. Clustering has been performed using affinity propagation. ANN algorithm has been used. Certain parameters such as accuracy, sensitivity, specificity, Recall and F-score has been calculated. The results demonstrate that the proposed diagnostic approach may properly predict the risk level of heart diseases.