MODELLING CRIME RATES IN INDONESIA USING TRUNCATED SPLINE ESTIMATOR

Abstract

Criminal acts are actions that violate the law and can arise from various factors such as emotions, psychological pressure, and others. Crime rate is a number that indicates the level of crime vulnerability in a certain area at a certain time. Higher crime rates correspond to increased vulnerability in an area and vice versa. Among various forms of criminal acts, the number of criminal acts and narcotics crimes in Indonesia tends to increase in 2020 and 2021. The aim of the research is to identify the characteristics of crime rate data based on the number of decency and narcotics incidents in Indonesia using a nonparametric regression approach. This research uses a nonparametric regression method spline truncated, and linear regression as a comparison. It was found that West Papua Province has the highest crime rate, based on a comparison between the linear regression model and the truncated spline nonparametric regression model, it can be concluded that the best model is the truncated spline nonparametric regression model with a Generalized Cross Validation (GCV) of 2468.487 and a coefficient of determination of 0.7389091, indicating that approximately 73% of the variability of the dependent variable can be explained by the independent variables included in the model.