Fatigue detection of vehicular driver through skin conductance, pulse oximetry and respiration: A random forest classifier

Since the 1990s, fatigue driving has been one of the two main causes of traffic accidents. In order to reduce traffic accidents, many models are applied in the detection of the driver's fatigue. This paper introduces a new model for detecting fatigue driving. This model employs the Hilbert-Huang transforms and the Random Forest Classifier algorithm for the analysis of the three factors, namely, skin conductance, oximetry pulse and respiration signals, and uses these parameters Accurate Rate, MSE, ROC, F1_score, Precision, and Recall for the evaluation of the model. It turns out that the new model improves the prediction accuracy of fatigue driving in comparison with MLP and SVC.