Applying Machine Learning to Predict Intercity Bus Ticket Prices During the Holiday Season

Abstract

The construction of the Trans Java Toll Road has intensified competition in the intercity bus (AKAP) industry, prompting bus operators (PO) to formulate strategies to maximize revenue. During holiday and festive seasons, operators typically raise ticket prices based on competitor pricing. However, other critical factors are often overlooked, such as fleet maintenance costs, travel operational expenses (fuel, tolls, driver commissions, meal allowances, and incentives), as well as agent commissions and passenger meals, all of which tend to increase—sometimes drastically—especially when fuel prices rise. This study aims to analyze passenger behavior, route potential, conduct preventive maintenance planning, and evaluate the alignment between ticket pricing, contribution margin, and cost ratio using a machine learning approach with the Random Forest Regression algorithm. The machine learning model was developed using data on route masters, fleet details, travel costs, ticket prices, holidays, and transaction records of revenue and operational expenses. The model achieved an RMSE of 3%, R² score of 77%, MAPE of 9.56%, and accuracy of 90.44%. It was then implemented in a web-based interface using Flask and ASP.NET to compare actual ticket prices with predicted prices.