SUMMARY
One of the problems in the distribution of electrical power is the identification of the location of the disturbance in the distribution network of pt pln. Several methods have been proposed to determine the location of the disturbance including the nearest neighbor method, impedance method and nave bayes method, in this study the author proposes the naïve bayes method. The naïve bayes method is a classification method based on simple probability and is designed to be used with the assumption that between one class and another class are not mutually dependent (independent). In the naïve bayes classification, the learning process is more focused on estimating probabilities. From the results of the experiment using 30 samples, the average error value for comparison of identification using the nave bayes method and the direct tracing method in identifying the location of disturbances in the buagan feeder was 0.86% with an average difference between the actual distance and the distance in the program of 30.9 meters. . The results of the comparative analysis using paired samples statistics showed no significant difference at 0.05. In the results of the comparative analysis using anova one way at 0.01 there is no significant difference. This study serves to narrow the direct line area for handling electrical disturbances on the 20 kv network, so that in the event of an electrical disturbance on the 20 kv network it does not take a long time in the process of identifying the location of the electrical disturbance.