IMPLEMENTATION OF K-NEAREST NEIGHBOR (KNN) AND SUPORT VECTOR MACHINE (SVM) FOR CLASIFICATION CARDIOVASCULAR DISEASE

WIJI LESTARI (Faculty of Computer Science, University of Duta Bangsa Surakarta)
SRI SUMARLINDA (Faculty of Computer Science, University of Duta Bangsa Surakarta)

Multicience
Vol. 2 No. 10 (2022): INTERNATIONAL JOURNAL OF MULTISCIENCE - JANUARY, 2022 EDITION

Data mining and machine learning are two tools that play an important role in the study of data analysis and decision systems. Classification is a function of data mining. In the classification function, sorting or mapping occurs based on the proximity or similarity of data attributes to the specified label. K-Nearest Neighbor (KNN) and Support Vector Machine (SVM). The KNN algorithm is a non-parametric method used for classification and regression. Cardiovascular disease prediction models with KNN and SVM algorithms are used to identify and predict cardiovascular disease. The KNN algorithm uses Euclidian distance for the prediction process of training data. The SVM algorithm uses a hyperplane for the data prediction training process. The dataset used is 400 with 7 attributes, namely age, gender, systolic blood pressure, cholesterol, talach, oldpeak and slope. The results of the implementation of the KNN and SVM algorithms produce performance with an accuracy of 75.75% on KNN and 76.00% on SVM. The precision value is 76.78% for KNN and 83.93% for SVM. Meanwhile, the recall resulted in 77.14% for KNN and 67.14% for SVM.