Artificial Neural Network Model Development to Predict Theft Types in Consideration of Environmental Factors

Abstract

Crime prediction research using AI has been actively conducted to predict potential crimes—generally, crime locations or time series flows. It is possible to predict these potential crimes in detail if crime characteristics, such as detailed techniques, targets, and environmental factors affecting the crime’s occurrence, are considered simultaneously. Therefore, this study aims to categorize theft by performing k-modes clustering using crime-related characteristics as variables and to propose an ANN model that predicts the derived categorizations. As the prediction of theft types allows people to estimate the features of the possibly most frequent thefts in random areas in advance, it enables the efficient deployment of police and the most appropriate tactical measures. Dongjak District was selected as the target area for analysis; thefts in the district showed four types of clusters. Environmental factors, representative elements affecting theft occurrence, were used as input data for a prediction model, while the factors affecting each cluster were derived through multiple linear regression analysis. Based on the results, input variables were selected for the ANN model training per cluster, and the model was implemented to predict theft type based on environmental factors. This study is significant for providing diversity to prediction methods using ANN.

Método de clustering e inteligencia artificial para clasificar y proyectar delitos violentos en Colombia

“…Por otra parte, estudios han utilizado los modelos de redes neuronales artificiales para predecir los tipos de robos considerando factores ambientales; sin embargo, explican que esto es posible solo si se integran las características y variables relacionadas con ese tipo de delito. Igualmente, resaltan la importancia de incursionar en áreas criminalísticas para proporcionar diversidad a los modelos de redes neuronales (Kwon et al, 2021), tal como se hizo en esta investigación.…”

An Integrated Graph Model for Spatial–Temporal Urban Crime Prediction Based on Attention Mechanism

“…In recent years, spatial-temporal crime prediction technology has been rapidly developed. With deriving data that include the crime incidents number, population density, weather variables, etc., spatial-temporal patterns of assault, robbery, theft, or other types of crimes can be predicted with the help of machine learning (especially deep learning) and other methods [1][2][3][4][5][6][7][8]. It provides references and predictions about when and where the crime hotspot would be to the police in advance, so it contributes to crime prevention as well as better police resources allocations.…”