Accident Prediction using Machine Learning

Accidents are major cause in cities. Various factors contribute to accident. The accident could be because of the driver, driving at a high speed or the driver condition. But other factors like weather, visibility, Road Condition, etc also contributes to making of accidents. The paper leverages established various machine learning techniques implemented on a dataset, which contains over 3000 data. The data is trained and tested with various techniques and the accuracy is tested for each of them. Also, an additional dataset with just 300 values was also trained using the same technique to see for the differences and similarity in the accuracy of the model. The model’s output gives the estimate of the factor which contributes largely towards the making of accident. Additionally, it also shows the day and the time, when accidents have occurred the most.