Dynamic Ticket Price Prediction Models for Spanish Renfe Railways: A Comparative Analysis of Machine Learning Algorithms

With the increasing demand for energy-efficient transportation solutions worldwide, there has been a significant reliance on trains as a preferred mode of travel. However, train ticket prices often fluctuate dynamically based on factors such as demand and time, leading to significant variations in pricing. This variability adversely affects passengers seeking affordable prices. To address this issue, this paper proposes a machine learning model for dynamic price prediction (DPP) utilizing a dataset derived from the Renfe AVE (Spanish High-Speed Train Service) ticket monitoring system. The study employs several algorithms, including multi-linear regression (MLR), polynomial regression (PR), K-nearest neighbors (KNN), decision trees (DT), random forests (RF), AdaBoost (ADB), and XGBoost (XGB). The primary objective of this research is to compare the effectiveness of different prediction methods and identify the approach that yields the most accurate results.