Prediction of Payment Modes for Online Taxi Users Using Machine Learning

It is critical to comprehend customer payment preferences for online taxi service providers in the digital age to give payment options that live up to client expectations. The variables affecting online taxi passengers' choice of payment method are frequently difficult for stakeholders to pinpoint. The purpose of this study is to use the random forest algorithm to forecast the preferred payment methods for online taxi users. The random forest approach was chosen due to its ability to handle complex information and generate accurate forecasts. An online taxi-user dataset with the labels “cash” and “digital” was employed throughout the model development stage. The random forest model is subsequently trained by taking into account the characteristics that affect the prediction of payment methods, such as income, online taxi brand, credit card, e-wallet, and account ownership. The created predictive model was tested on the dataset using an experiment employing the bootstrapping method after going through the training phase. An accuracy rate of 96% in the experiment demonstrates how well the random forest model predicts the mode of payment for online taxi consumers. In addition, high values of precision, recall, and Fl-score were also obtained, indicating the reliability of the model in correctly classifying payment modes.