A fatigue driving detection algorithm based on weighted D-S evidence theory

Fatigue driving detection is of great significance to public traffic safety. Traditional fatigue driving detection algorithm pursues the minimum classification error rate and ignores the loss caused by different misclassification. However, fatigue driving detection algorithm should pay more attention to the high cost misclassification samples to minimize the overall loss of the learning model. In order to solve this problem, the cost-sensitive learning and Dempster-Shafer evidence theory (D-S evidence theory) are combined to obtain a robust and good performance ensemble classifier.