Forecasting the Number of Road Accidents in Poland Depending on the Day of the Week using Neural Networks

Autor: Gorzelanczyk Piotr, Tylicki Henryk
Jazyk: angličtina
Rok vydání: 2023
Předmět:
Zdroj: Logi, Vol 14, Iss 1, Pp 35-42 (2023)
Druh dokumentu: article
ISSN: 2336-3037
DOI: 10.2478/logi-2023-0004
Popis: The number of road accidents in the world is decreasing year by year. This number has been affected by the pandemic in recent years but is still very high. Therefore, it is necessary to take all possible measures and steps to reduce this number. The objective of the article is to forecast the number of road accidents in Poland depending on the day of the week. For this purpose, the annual data on the number of road accidents in Poland broken down by days of the week were analyzed, and a forecast for the years 2022-2040 was prepared on the basis of the police statistics. The forecast was made using selected models of neural networks. The research results show that a decrease in in the number of road accidents can be expected. However, the obtained results could be affected by the selected number of random samples (training, testing and validation).
Databáze: Directory of Open Access Journals