Abstract: Indian railways are one of the largest railway networks which are managed by the Indian government. Indian trains are overcrowded most of the time due to huge demand and very less supply. Train delays are mainly caused by train speed restrictions imposed for safety reasons which can disrupt the train schedules for the connecting journeys. Real-world data was obtained to perform the research. The data was cleaned, pre-processed, and dimensionality reduction was done. Further, five machine learning models for classification were applied. The highest prediction accuracy was obtained by the Random Forest model.

Keywords: Machine Learning, Classification Models, Train Delay, Real-World Data, Delay Prediction, Transit Delays

| DOI: 10.17148/IARJSET.2021.8325