Comparison of Naïve Bayes and KNN Algorithms to understand Hepatitis

The heart is the most important organ for humans. The liver functions to neutralize toxins that are in the blood and regulate the composition of blood that contains fat, protein, sugar and other substances. The Hepatitis is the disease that attacks the liver caused by a virus. Hepatitis can be known by holding a laboratory test on the blood. The development of technology and information on hepatitis can be known by the classification and prediction methods. The purpose of this study was to improve the accuracy of the classification of naïve Bayes and KNN algorithms by taking public data from the UCI Repository with total of 155 data, having 19 attributes owned such as Age, Gender, Steroids, Antivirus, Fatigue, Malaise, Anorexia, Big Heart, Heart Company, Spleen, Spiders, Ascites, Varicose, Bilirubin, Alk Phosphate, Shot, Albumin, Protime, Histology, and Class (predictive attribute). Experiments use the confusion matrix to determine the value of accuracy, precision, and recall. The results obtained in experiments using Naïve Bayes algorithm are the level of accuracy of 74.19% and the average level of error 25.81% higher than the K-Nearest Neighbor algorithm the average value is 54.84% and the level of value an average error of 45.18%. From the results obtained that the K-Nearest Neighbor algorithm increases the value of accuracy and the average value of errors from previous studies.