Fatality Prediction for Motor Vehicle Collisions: Mining Big Data Using Deep Learning and Ensemble Methods

Motor vehicle crashes are one of the most common causes fatalities on roads. Real-time severity prediction such may contribute towards reducing rate fatality. In this study, fundamental goal is to develop machine learning models that predict whether outcome a collision will be fatal or not. A Canadian road crash dataset containing 5.8 million records utilized in research. ensemble have been developed using majority and soft voting address class imbalance dataset. The accuracy approximately 75% achieved Convolutional Neural Networks. Moreover, comprehensive analysis attributes important distinguishing between vs. non-fatal motor collisions has presented paper. In-depth information content reveals factors model. These include roadway characteristics weather conditions at time crash, type, when happen, user their position, any safety device used, status traffic control. With real-time data based conditions, an automated warning system can potentially utilizing model employed study.