Predict the Number of Traffic Accidents in Turkey by Using Machine Learning Techniques and Python Tools

  • MUSTAFA AL-ASADI Selçuk University
  • Şakir TAŞDEMİR Selcuk University Department of Computer Engineering
  • Humar KARAMANLI ÖRNEK Selcuk University Department of Computer Engineering
Keywords: Machine learning, Traffic Accidents Prediction, Mean Absolute Error, Python

Abstract

Traffic accidents have undergone rapid growth in recent years worldwide, causing great life and property losses. Therefore, predicting traffic accidents is a crucial problem for improving transportation and public safety. Machine learning is one branch of artificial intelligence that can extract information from data and use statistical methods to predict values. In this paper, three machine-learning techniques were applied to estimate the number of traffic accidents (with death or injury) in Turkey until the year 2029, which are: Linear Regression (LR), Decision Trees (DT), and Random Forest (RF). These techniques were tested using a real data set obtained from the website: http://www.tuik.gov.tr and we got the result that Linear Regression (LR) has the best performance. This result indicates the approach's predictive superiority in predicting road accidents. As a result, the study will assist road transport and insurance agencies in developing road safety strategies.

Published
30-12-2022
How to Cite
AL-ASADI, M., TAŞDEMİR, Şakir, & ÖRNEK, H. (2022). Predict the Number of Traffic Accidents in Turkey by Using Machine Learning Techniques and Python Tools. Artificial Intelligence Studies, 5(2), 35-46. Retrieved from https://www.aistudies.org/index.php/ais/article/view/56
Section
Articles