PENERAPAN METODE NAIVE BAYES UNTUK MEMPREDIKSI STATUS KEHADIRAN MASYARAKAT DALAM PEMILIHAN GUBERNUR

Abstract

General Election (PEMILU) is an important political event to determine the leader in a democratic country, where the process of people's interests, which is then formulated in various forms of policy. Various problems related to Permanent Voter Data (DPT) including the KPU difficult to get the NIK of people in the Correctional Institution, novice voters who do not have an Identity Card (KTP), voters who do not have a resident identity. there are people who should have had criteria as voters but were not registered, while people who have died are still registered. In addition, there are voters who are registered at more than one polling station (TPS) or are called multiple voters and voters who are difficult to find so that the KPU must visit houses as stipulated in the legislation. This study aims to facilitate the classification of community attendance status in the governor election using the Naive Bayes method, the data used are Permanent Voter List (DPT) data obtained from the Tanjungpinang City General Election Commission. The input variables used are gender, marital status, address, distance home, job, status of existence, ownership of ID card. While the output variables used were present and absent. The data used were 250 data ,70% for training data and 30% for testing data. The data tested were 76 data with an accuracy value of 78,95%.