Drowsiness Detection of Driver using Novel Random Forest Classifier and Logistic Regression Classifier with Improved Accuracy

Abstract

Aim: The aim is to detect the driver drowsiness using Logistic Regression Classifier method in Comparison with Novel Random Forest Classifier. Materials and Methods: Two groups such as logistic regression Classifier and the novel Random Forest Classifier are applied. Each group has a sample size of 10. SPSS was used for predicting significance value of data set considering G-power value as 80%. Using different sample sizes, their accuracies are compared to each other. Result: Novel Random forest classifier provides a higher accuracy of 63.20% when compared to Logistic Regression Classifier with accuracy 60.00% in predicting stock driver drowsiness detection. There is a significant difference between two groups with a significance value of 0.001 (p<0.001). Conclusion: The results show that the Novel Random Forest classifier detects driver drowsiness better than the logistic Regression Classifier algorithm. It can also be thought of as a superior choice for classifying driver drowsiness.