India operates the fourth largest railway network in the world, transporting over 8 billion passengers annually. Despite this extensive network, passengers frequently encounter delays, causing significant inconvenience. Given the vast number of passengers, improving the punctuality of trains in India is crucial. This study aims to model the delays of selected trains to understand their patterns over time. Train delay data, collected at various points over a period, has been used for this analysis. Researchers have developed functions to extract and process this delay data from HTML content obtained from a list of URLs. The study focuses on delays in the arrival of 20 trains at their destinations, employing various tools and machine learning algorithms, particularly regression-based methods, to model this data. Findings indicate that the Extra Trees Regression model outperforms other regression models in predicting train delays.