Abstract:

In the urban transportation sector, accurately predicting the hourly boarding demand of bus passengers is crucial for optimizing resource allocation, improving service quality, and enhancing passenger satisfaction. This project addresses the challenge of forecasting bus passenger demand using imbalanced records obtained from smart-card data, which are often skewed and sparse. We propose a deep learning approach that leverages advanced techniques to handle data imbalance and extract meaningful patterns from the complex and voluminous smart-card records. Our methodology involves pre-processing the raw data to address imbalances and employing a deep neural network model to predict hourly boarding demand. The model is trained on historical smart-card data, which includes various features such as time of day, day of the week, weather conditions, and geographical information. By incorporating these factors, the model can accurately capture the temporal and spatial dynamics of passenger demand. The proposed deep learning model demonstrates superior performance compared to traditional forecasting methods, achieving higher accuracy and robustness in predicting hourly passenger demand. The results of this study provide valuable insights for urban transit authorities to enhance their operational efficiency and decision-making processes. This project contributes to the field of intelligent transportation systems by presenting a scalable and effective solution for demand forecasting in public transit, ultimately leading to more efficient and passenger-friendly bus services.

Description:

.