Prediction of Heart Disease using Different KNN Classifier

摘要

Recently, Machine learning classification algorithms are playing a vital role in analysing various data available in cloud storage and websites. In this paper, the heart disease dataset is considered and the results are predicted by using various version of the KNN classifier in MATLAB. The comparisons of the performance of all these algorithms have been evaluated for accuracy, misclassification rate. The distance metric and distance weight used by various KNN algorithms is also shown. The experiment has done with the PCA option enabled. The optimized KNN was best with an accuracy of 69% with a prediction speed of ~5600 obs/sec.