Urban traffic accident prediction research based on meteorological data

With the increase in the number of all megacities and large cities, urban traffic has become an important issue in urban life. Urban traffic accidents are occurring more and more frequently, which has a negative impact on the economy and safety. The existing research shows that sunlight intensity or rainy days can increase the probability of accidents in urban traffic. This paper proposes multiple machine learning methods to compare the accuracy of different machine learning algorithms for New York traffic accident prediction on weather data from 2015 to 2020. The experimental results show that both OVR and OVO methods have high accuracy except for the lower accuracy in 2020.