Region-level demand forecasting is an essential task in ridehailing services. Accurate ride-hailing demand forecasting can guide vehicle dispatching, improve vehicle utilization, reduce the wait-time, and mitigate traffic congestion. This task is challenging due to the complicated spatiotemporal dependencies among regions. Existing approaches mainly focus on modeling the Euclidean correlations among spatially adjacent regions while we observe that non-Euclidean pair-wise correlations among possibly distant regions are also critical for accurate forecasting. In this paper, we propose the spatiotemporal multi-graph convolution network (ST-MGCN), a novel deep learning model for ride-hailing demand forecasting. We first encode the non-Euclidean pair-wise correlations among regions into multiple graphs and then explicitly model these correlations using multi-graph convolution. To utilize the global contextual information in modeling the temporal correlation, we further propose contextual gated recurrent neural network which augments recurrent neural network with a contextual-aware gating mechanism to re-weights different historical observations. We evaluate the proposed model on two real-world large scale ride-hailing demand datasets and observe consistent improvement of more than 10% over stateof-the-art baselines.

safety issues of organic electrolyte, and lithium dendrites. [3][4][5][6] Alternatively, rechargeable aqueous zinc-ion batteries (ZIBs) are increasingly attracting attention owing to their straightforward processing, high capacity of Zn metal anode (820 mAh g −1 ), [7] excellent safety, cost effectiveness (US$65 kWh −1 ), [8] higher ionic conductivity of aqueous electrolytes (≈1 S cm −1 ) than non-aqueous electrolytes (1-10 mS cm −1 ), [8] and two-electron transfer mechanism. [9] However, previous works of ZIBs based on the alkaline electrolytes still retain formidable challenges, such as inferior cyclability (≤1000 cycles), low coulombic efficiency (≤90%), and limited capacity (≤400 mAh g −1 at 0.1 A g −1 ), [1] mainly caused by the byproducts and unstable cathodes.To solve the above issues, much attention has been devoted to exploring highcapacity, electron-ion conductive, and structurally stable cathode materials, such as manganese oxides, [10][11][12] Prussian blue analogues (e.g., Zn 3 [Fe(CN) 6 ] 2 ), [13] NASICON-type materials (e.g., Na 3 V 2 (PO 4 ) 3 ), [14] transition metal sulfide (e.g., VS 2, Mo 6 S 2 ), [15,16] polymer (e.g., pyrene-4,5,9,10-tetraone, [17] polyaniline, [18] ), and vanadium-based compounds. [19][20][21] Recently, layered vanadium-based materials, including Zn 0.25 V 2 O 5 ·nH 2 O, [8] [9] and zinc orthovanadate, [24] have been recognized as a very promising class for high-safety aqueous ZIBs with respectable Zn-storage capacity of 300-400 mAh g −1 and cyclability (>1000 cycles), [25] originating from multivalence nature of vanadium cations and superior stability of layered structure. [22,26,27] It is noteworthy that all of the layered vanadium-based materials reported are composed of crystalline frameworks with doped metal ion or structural water, which neither sufficiently accommodates the structural strain during the (de)intercalation process nor provides ample active sites for efficient charge storage. [28,29] In a sharp contrast, amorphous frameworks can not only provide more exposed ion channels, accelerate rapid charge transfer across the electrode/electrolyte interface, and further facilitate fast ion intercalation, but also offer low internal energy and outstanding chemical stability. [30,31] Moreover, 2D heterostructures by alternating stacking of two type different 2D nanosheets produce Rechargeable aqueous zinc-ion batteries (ZIBs) are appealing due to their high safety, zinc abundance, and low cost. However, developing suitable cathode materials remains a great challenge. Herein, a novel 2D heterostructure of ultrathin amorphous vanadium pentoxide uniformly grown on graphene (A-V 2 O 5 /G) with a very short ion diffusion pathway, abundant active sites, high electrical conductivity, and exceptional structural stability, is demonstrated for highly reversible aqueous ZIBs (A-V 2 O 5 /G-ZIBs), coupling with unprecedented high capacity, rate capability, long-term cyclability, and excellent safety. As a result, 2D A-V 2 O 5 /G heterostructures for stacked ZIBs at 0.1 A g −1 dis...