Forecasting the Number of Passengers in Rail System by Deep Learning Algorithms

Public transportation has emerged as an important factor in selecting and developing urban centers, including diverse sectors, such as commerce, social engagement, education, and healthcare. The relationship between transportation and urban development underlines the crucial role of transportation in cities. The continuous increase in population density has led to a rapid increase in the number of urban passengers, thereby intensifying the complexity of public transportation networks. Consequently, developing strategic short-term and long-term transportation plans has become an essential task. In this study, we developed a prediction model for estimating the number of passengers in subways. In this manner, we used Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM) algorithms, which are reported to have high predictive performance in time series. To test the developed model, we used New York Subway data. The results show that the RNN algorithm has a high prediction performance in estimating the number of passengers.