Prediction of traffic accident numbers using time series analysis models: The case of the Western Black Sea Region

This thesis study aims to estimate the number of fatal and injured traffic accidents that occur in the Gendarmerie area of responsibility in the Western Black Sea Region with time series models. In the study, statistical data from April 2019 to December 2023 on the official website of the Gendarmerie General Command were used. These data include traffic accidents that occurred in the provinces of Bartın, Bolu, Düzce, Karabük, Kastamonu, Sinop and Zonguldak. In the study, the number of future accidents was estimated using the autoregressive integrated moving averages (ARIMA) method. In this context, data analysis and modeling were carried out with Minitab, R Studio and EViews programs. The results of the study were supported by trend analysis, seasonality analysis and forecast modeling. According to the findings of the thesis, estimating the number of traffic accidents will contribute to decision-making managers making more informed and effective decisions on issues such as personnel assignment and economic planning. In addition, in similar studies to be carried out in the future, which concern academic and public institutions, the estimation results obtained will also help to take the necessary measures to prevent traffic accidents, which are important for public health and the state economy.