Prediksi Ancaman Yang Dihadapi KORPSBRIMOB Polri Dengan Menggunakan Metode K-Nearest Neighbor Classifier Machine Learning dan Naive Bayes

Abstract

This research specifically applies two classification algorithms—K-Nearest Neighbors (K-NN) and Naïve Bayes—to assess and predict social disturbances. The empirical findings indicate that, in the context of social factors, the Naïve Bayes algorithm achieved a classification accuracy of 93.91%, surpassing K-NN's 73.33%. These results suggest a significant likelihood of social disturbances during the 2025–2026 period. Such disturbances, often involving the Police Brimob Corps, are predominantly driven by social tensions, including intergroup conflict, economic grievances, dissatisfaction with governance, and identity-based strife.The prevalence of socially-rooted unrest underscores the need for multifaceted intervention strategies. As a specialized unit within Indonesia’s national security apparatus, the Brimob Corps is positioned to mitigate these threats through integrated approaches ranging from mass security and enforcement operations to intelligence-led dialogue and preventive engagement. This highlights the vital role of predictive analytics in shaping proactive and adaptive security policies.