Forecasting Criminal Offenses Against Persons Using Time Series Models: A Case Study of Mwanza Region

The study aimed to determine the best statistical forecasting model that best fit the criminal Offenses against persons in Mwanza. The study used yearly time series data of criminal Offenses against persons from the Tanzania Police Force for the period 1960 to 2018. The findings show that yearly average, the criminal Offenses against persons reported in the Mwanza region was 269, the maximum was 867 and the minimum was 44. The criminal Offenses against persons reported in Mwanza have an upward trend which is inconsistent from year to year. The RSME, MAPE, AIC and BIC were used to assess the forecasting accuracy of the models. The models included the ARIMA model, SMA and SES. On the basis of RSME, MAPE, AIC and BIC the results showed that ARIMA (1, 1, 1) model is the best model that fits the criminal Offenses against persons in the Mwanza region and therefore suitable for forecasting criminal Offenses. The model that has been developed is a useful tool for producing reasonably reliable forecasts of criminal Offenses against persons in future years. These forecasts can provide guidelines in understanding whether the criminal offence against persons rising or falling.