Improved Detection of Truck Failure Due to Air Pressure System by Novel Xgboost Algorithm over Decision Tree Algorithm

Abstract

Aim: To improve the detection of a truck failure due to an Air Pressure System (APS) using Machine learning algorithms and improve the accuracy of APS truck failure using the  Novel XGBoost algorithm. Materials and Methods: The Novel XGBoost algorithm is used for the prediction of truck failure due to the Air Pressure System. The dataset has a total sample size of 10 for each group and  SPSS package utilized for the performance analysis of accuracy in detecting the air pressure system.  To improve the accuracy to detect truck failure due to an Air Pressure System using Novel XGBoost is proposed and compared with the decision tree algorithm. Results and Discussion: Test results prove that Novel XGBoost has an average accuracy of  98.24% which is better than the Decision Tree has an average accuracy of 96.62%. The analysis of the training dataset and testing dataset has been performed successfully using SPSS and acquired 98.24% accuracy for predicting the truck failure due to the Air pressure system. With the level of significance (p<0.05) the resultant data depicts the reliability of independent sample tests. Conclusion: The overall process of prediction of accuracy using the Novel XGBoost algorithm gave significantly better results compared to the Decision Tree algorithm.