A Model for Estimation of Crime Prediction in Katagum Zone of Bauchi State, Nigeria

ABSTRACT:

The wave of criminal activities in Nigeria is increasing daily and the need to reduce it to a minimal level is the cardinal responsibility of government through its well-established security agencies, such as, the Nigeria Police Force. This study examines a model for estimation of crime prediction in Katagum zone ofBauchi State, Nigeria. The datasets used for the study were collected from Katagum Police Area Command of the state. Autoregressive Integrated Moving Average (ARIMA) model was employed for the estimation and ARIMA (1,1,2) was found suitable for the prediction of crime rate. The model was implemented in Python. The datasets were divided into 70% training set and 30% testing set. Mean Absolute Percentage Error (MAPE) was adopted for metrics accuracy and computed to be 11.4 % indicating that the model has 88.6% accuracy. The results showed that in January 2020, total crime cases were 43±8. Meanwhile, the maxima crime cases occurred in December with predicted rate of 53±12 cases. Besides, the results also revealed that there was steady increase in year 2021 with January having records of 53±12 crime cases. The maximum crime rate was also in December, 2021 with crime predicted of 55±19 cases. From these findings, the crime cases occur minimally in the beginning of the year, that is, January, and steadily increases throughout year, maximally in the month December. The study recommends that more police officers should be deployed to the Katagum Police Area Command for proactive measures, especially in the last month of the year.

KEYWORDS: ARIMA Model, Crime rate, Katagum Zone, Prediction, Security Agencies