Implementasi Model Local Outlier Factor (LOF) dalam Deteksi Anomali pada Data Pemilih KPU Kabupaten Sukabumi

Elections are one of the most significant political activities in the life of a nation, requiring accurate and reliable voter data. Inaccurate or unreliable voter data can lead to various issues, such as election fraud. One of the main causes of inaccurate voter data is the presence of anomalies, which are data points that do not reflect actual conditions. Anomalies in voter data can arise from several factors, including data entry errors, fraud, or system failures. To detect anomalies in voter data, various Machine Learning methods can be employed, one of which is the Local Outlier Factor (LOF) method. LOF identifies anomalies by measuring the distance between data points and their nearest neighbors. This study demonstrates that the LOF method can effectively detect anomalies in the voter data of the Sukabumi Regency General Election Commission, particularly in numerical datasets such as age, with a very high accuracy rate of 99.98%.

Keywords: General Election, Voter Data, Anomaly, Local Outlier Factor (LOF), Machine Learning, Sukabumi Regency Election Commission