ORIGIN-DESTINATION PASSENGER FLOW PREDICTION IN METROS USING ADAPTIVE FEATURE FUSION

In order for metro systems to function optimally and deliver exceptional service, accurate customer volume forecasts are necessary. A groundbreaking method called Adaptive Feature Fusion (AFF) is described in this article. In order to enhance the forecasting of origin-destination (OD) flows, it actively incorporates external, temporal, and geographical data. In contrast to conventional models, which evaluate all characteristics consistently, AFF prioritizes certain attributes depending on context. This makes it possible to generate more precise estimations that can be adjusted according to the actions of passengers. After testing AFF with real-world metro data, researchers found it to be more accurate and dependable than competing approaches. Less congestion and an improved experience for riders are outcomes of this research's contribution to smarter urban transportation planning.