Prediction of Crime Rate in Banjarmasin City Using RNN-GRU Model

Abstract

Crime is a crime that violates the law and social norms so that it can harm society. Every year there is an increase in criminal cases. The rise of various criminal acts caused disturbances to the community's comfort and the surrounding environment, especially in Banjarmasin City, the capital of South Kalimantan Province in Indonesia. The economic status of a region, such as the inflation rate and the local population's lack of purchasing power, can contribute to crime. This study proposes a model to predict the crime rate in Banjarmasin using the Recurrent Neural Network (RNN) with the Gated Recurrent Unit (GRU) architecture, taking inflation rate and discretionary income into consideration. This study utilizes data on criminal offences handled by the Banjarmasin District Court and data on inflation and the cost of staple foods in the Banjarmasin City markets. We evaluate the model by using RMSE and R-Squared. The results showed that the GRU-RNN model showed promising results with an R-Squared value of 0.84 and an RMSE value of 2.21.