Covid-19 classification using K-nearest neighbor with various distances Available to Purchase

K-Nearest Neighbor or also known as KNN is one of classification method in data mining. KNN is used to classify data with label or supervised data. The KNN algorithm classifies unlabeled test samples based on the majority of similar samples among the closest neighbors to the test sample. KNN algorithm requires distance metrics to classify, there are lots of distance metrics. Continuely, Euclidean distance is often used in clustering or classification, in this research comparatively use Covid-19 symptoms dataset then we compare KNN with Euclidean distance to KNN with other distances i.e., Manhattan, Chebyshev, Bray-Curtis, Canberra, and Cosine use the classification’s accuracy, precision, sensitivity/recall, F-1 score and time to see wich one is better for the dataset. This research produce the accuracy, precision, recall rate and F-1 Score respectively are same with value is 0.98 for Euclidean, Manhattan. The time generated by Chebyshev is faster than other distances with the computation time is 36.4 seconds.