Criminal Cases Forecasting Model using A New Intelligent Hybrid Artificial Neural Network with Cuckoo Search Algorithm

摘要

Criminal cases are social problems that concern the public order and good morals of the citizens as a whole. Research in criminology has recently focused on finding the root causes of criminal cases in order to identify the factors accelerating crimes and to find preventive methods. Thus, the study on the factors relating to the occurrence of crimes would benefit policy makers in dealing with crimes. This research focuses on a forecasting model of the criminal related factors in the Upper Northeastern Provinces of Thailand by using a New Intelligent Hybrid Artificial Neural Network (NIHANN) with a Cuckoo Search Algorithm (CS). The new CS-NIHANN algorithm, which is a combination of the CS with the NIHANN, is used in this research. In addition, the research selected relevant parameters that related to criminal cases in the Upper Northeastern Provinces of Thailand. The comparison of performances of the forecasting model for historical forecast part was calculated by Mean Absolute Percentage Error (MAPE), Root Mean Squared Error (RMSE) and Correlation Coefficient (R). The result shows that the CS-NIHANN algorithm trains fast, can obtain the global optimal solution, and has a good generalization performance for Multiple Input Single Output (MISO) and Multiple Input Multiple Output (MIMO). Furthermore, the result shows that the CS-NIHANN algorithm has a faster convergence than the CS-ANN algorithm in calculating an optimal solution to the criminal cases.