Utilizing a Five-year Bayesian Predictive Analysis in Crime Rate

As the crime rate has continuously had its share of rise and fall for the past years in the country, the study aims to investigate, with the utilization of Bayesian Statistics, a probable predictive model for the determination of crime rates for the next five years. With data coming from the Crime Recording and Analysis System (CRAS) from the years 2020-2024, the study made use of statistical modeling, which allowed the determination and comparison of various models. The forecast determined a steady decline of the crime rate from 38 per 100,000 total population to 21.93 in the year 2029. With the determination of tested models, it was concluded that the 2024 model had the strongest predictive power based on the process of Bayesian model comparison, having the highest posterior probability and an acceptable variance, hence justifying the decline of the crime rate for the next five years. The findings provide valuable insights for the future strategizing of the Philippine National Police, as well as allowing the formulation of data-driven policies and better resource allocation. Keywords: Crime Rate, Forecast, Bayesian Statistics, Predictive Analysis, Prognostic