Equity-Aware Cross-Graph Interactive Reinforcement Learning for Bike Station Network Expansion

Xi Yang, Suining He, Mahan Tabatabaie

Abstract: Thanks to advances in the urban big data, the bike sharing, especially station-based bike sharing, has emerged as the important first-/last-mile connectivities in many smart cities. Bike station network (BSN) expansion recommendation, i.e., recommending placement locations of new stations, is essential for satisfying local mobility demands, enhancing the BSN service quality, and may significantly affect the resource fairness and accessibility of different communities in the neighborhood. Furthermore, the dynamic and complex urban mobility environments make the station placement highly challenging to satisfy the mobility needs.To ease and facilitate the urban planning with awareness of mobility equity, we have designed and proposed CGIRL, a novel equity-aware Cross-Graph Interactive Reinforcement Learning approach for BSN expansion recommendation. Specifically, we have designed a novel reward function within our actor-critic reinforcement learning approach, jointly accounting for the local mobility, bike resource distribution equity, and accessibility of different socioeconomic groups to the expanded stations. To capture the policy of station decisions from the BSN deployment, we integrate the location graph's mobility and equity correlations across the city regions as a graph network, and design a novel cross-graph interaction network with embedding attention and sequential dependency that adaptively captures and interactively differentiates the correlations within station placement. Our extensive experimental studies upon a total of 393 (111 new) bike stations from New York City (NYC), Washington D.C. (DC), and Chicago have validated the effectiveness of CGIRL in the equity-aware BSN expansion recommendation.