Forecasting road accidents in Poland and Croatia using artificial intelligence

Road traffic accidents remain a critical issue in Poland and Croatia, even though recent years have seen a decline in their numbers. The COVID-19 pandemic temporarily reduced mobility, but the persistent occurrence of accidents highlights the need for advanced predictive measures to improve road safety. This research leverages artificial intelligence, specifically neural networks, to predict the number of road accidents in Poland and Croatia from 2024 to 2030. Historical accident data were sourced from official records, including the Polish Police and Eurostat, and analyzed using multilayer perceptron (MLP) models. The results indicate a trend toward stabilization in accident rates, influenced by factors such as infrastructure development and rising vehicle ownership. However, the accuracy of these forecasts is constrained by the inherent variability in the data.