Public Transport Delay Prediction Using Machine Learning Techniques

Abstract

Delays in public transit have a substantial impact on urban mobility, resulting in passenger annoyance and decreased system reliability. The majority of current delay estimate techniques rely on set timetables and past averages, which are unable to adjust to changing traffic circumstances. In order to successfully understand delay patterns, this research suggests a machine learning-based public transportation delay prediction system that makes use of historical and traffic-related data. For precise delay prediction, a Random Forest model is trained using the preprocessed data. Users can enter travel-related data and get information about anticipated delays as well as explanations. The suggested approach increases forecast accuracy, facilitates better trip planning, and raises the general effectiveness and dependability of public transportation systems by adjusting to shifting traffic patterns.