Prediction of Heart Disease Using K-Nearest Neighbour Algorithm in Comparison with Support Vector Machine Algorithm

To classify the heart diseases according to the categorial identification by using the symptoms and the data of the Patient. Procedures and Materials: To increase the accuracy of K-nearest neighbors in predicting heart illnesses over Support Vector Machines. The proposed methodology evaluates a total of 10 samples in each of two groups to arrive at a significance value of 0.001. Results: The sample estimates were performed with the GPower statistical programme and yielded an actual power of 80%, an effect size of 0.589345, and a standard deviation of.76485. In diagnosing cardiac disease, the K-nearest neighbor classifier has a higher accuracy (87.1%) than the SVM classifier (81.6%). Conclusion: When compared to the Support Vector Machine, the K-nearest neighbor method performs much better in the prediction of heart disease.