Comparison of Naive Bayes and K-Nearest Neighbor methods to predict divorce issues

Abstract

Divorce in Cimahi every year continues to increase, each month to reach an average of 800 divorce cases, from the case of 75 cases of 75% household divorce cases, while the rest of the other cases, such as marriage isbat and inheritance. Based on the problem there is a need for prediction to find out how much divorce in each month. One of the techniques used to find divorce is by doing data processing to predict the occurrence of a divorce that is by using data mining techniques such as Naive Bayes algorithm and K-Nearest Neighbor. This algorithm has a high degree of accuracy in predicting. The best level of accuracy between the two algorithms can be determined by comparison. Comparison of algorithm aims to get the algorithm that is considered the fastest and accurate to make a prediction of a problem. Result of comparison of Naive Bayes and K-Nearest Neighbor algorithm can be concluded that Naive Bayes algorithm yield 72,5% accuracy and K-Nearest Neigbor algorithm yield 57,5% accuracy.