Predict GO-JEK Driver Income Level in the Bali Region using Decision Tree

Abstract

The work in the transportation industry is one of several that is impacted by information technology, which is now developing quickly. The emergence of information technology in the transportation industry led to the creation of GO-JEK. GO-JEK is growing in popularity with the general public because, in addition to offering a wide range of services, it also creates new jobs with more flexible working hours for the community. Of course, each driver's income from working as a GO-JEK driver is unique. However, a number of things can have an impact on this income. With a focus on the Bali region, this study tries to forecast income levels for drivers of GO-JEK. J48 algorithm is used to process the data, producing a decision tree with 19 rule models. An accuracy value of 77.61%, a precision value of 78.95%, a value recall of 81.08%, and an AUC value of 0.808 are obtained from the rule model test utilizing the confusion matrix and ROC/AUC curve. These results demonstrate how effective the decision tree and rule model are.