Modeling the Number of Road Accidents in Kenya Using Time Series Analysis: A SARIMA Approach
Oliver Siminyu Wamani *
Kaimosi Friends University, Kenya.
Michael N. Musyoki
Kaimosi Friends University, Kenya.
*Author to whom correspondence should be addressed.
Abstract
Road transport is a vital mode of mobility in Kenya. Despite existing research and numerous interventions, road accidents continue to pose a sig- nificant challenge. This study investigates the trend and future projection of road accidents using a Seasonal Autoregressive Integrated Moving Average (SARIMA) model. Secondary data from the National Transport and Safety Authority (NTSA) from 2019 to 2023 was analyzed using R (version 4.3.1). The optimal model, SARIMA (0,1,1)(1,0,0)[12], demonstrated high forecasting accuracy. The study provides forecasts for the next 24 months and offers insights for policymakers and safety planners to mitigate accident risks.
Keywords: Modelling, road accidents, SARIMA, NTSA, time series