Predicting Station-level Hourly Demands in a Large-scale Bike-sharing Network: A Graph Convolutional Neural Network Approach

Abstract This study proposes a novel Graph Convolutional Neural Network with Data-driven Graph Filter (GCNN-DDGF) model that can learn hidden heterogeneous pairwise correlations between stations to predict station-level hourly demand in a large-scale bike-sharing network. Two architectures of the GCNN-DDGF model are explored; GCNNreg-DDGF is a regular GCNN-DDGF model which contains the convolution and feedforward blocks, and GCNNrec-DDGF additionally contains a recurrent block from the Long Short-term Memory neural network architecture to capture temporal dependencies in the bike-sharing demand series. Furthermore, four types of GCNN models are proposed whose adjacency matrices are based on various bike-sharing system data, including Spatial Distance matrix (SD), Demand matrix (DE), Average Trip Duration matrix (ATD), and Demand Correlation matrix (DC). These six types of GCNN models and seven other benchmark models are built and compared on a Citi Bike dataset from New York City which includes 272 stations and over 28 million transactions from 2013 to 2016. Results show that the GCNNrec-DDGF performs the best in terms of the Root Mean Square Error, the Mean Absolute Error and the coefficient of determination (R2), followed by the GCNNreg-DDGF. They outperform the other models. Through a more detailed graph network analysis based on the learned DDGF, insights are obtained on the “black box” of the GCNN-DDGF model. It is found to capture some information similar to details embedded in the SD, DE and DC matrices. More importantly, it also uncovers hidden heterogeneous pairwise correlations between stations that are not revealed by any of those matrices.

[1]  Tianqi Chen,et al.  XGBoost: A Scalable Tree Boosting System , 2016, KDD.

[2]  Pierre Vandergheynst,et al.  Wavelets on Graphs via Spectral Graph Theory , 2009, ArXiv.

[3]  Zhenhua Zhang,et al.  Abnormal Spatial-Temporal Pattern Analysis for Niagara Frontier Border Wait Times , 2017, ArXiv.

[4]  Jiming Chen,et al.  Mobility Modeling and Prediction in Bike-Sharing Systems , 2016, MobiSys.

[5]  Xiqun Chen,et al.  Short-Term Forecasting of Passenger Demand under On-Demand Ride Services: A Spatio-Temporal Deep Learning Approach , 2017, ArXiv.

[6]  Michael Rabbat,et al.  How Does Land-Use and Urban Form Impact Bicycle Flows--Evidence from the Bicycle-Sharing System (BIXI) in Montreal , 2014 .

[7]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[8]  Peter Bühlmann.  Regression shrinkage and selection via the Lasso: a retrospective (Robert Tibshirani): Comments on the presentation , 2011 .

[9]  R. Alexander Rixey,et al.  Station-Level Forecasting of Bikesharing Ridership , 2013 .

[10]  Zhaohui Wu,et al.  Dynamic cluster-based over-demand prediction in bike sharing systems , 2016, UbiComp.

[11]  Max Welling,et al.  Semi-Supervised Classification with Graph Convolutional Networks , 2016, ICLR.

[12]  W. Y. Szeto,et al.  A modeling framework for the dynamic management of free-floating bike-sharing systems , 2018 .

[13]  W. Y. Szeto,et al.  A static free-floating bike repositioning problem with multiple heterogeneous vehicles, multiple depots, and multiple visits , 2018, Transportation Research Part C: Emerging Technologies.

[14]  Romain Giot,et al.  Predicting bikeshare system usage up to one day ahead , 2014, 2014 IEEE Symposium on Computational Intelligence in Vehicles and Transportation Systems (CIVTS).

[15]  Nando de Freitas,et al.  A Tutorial on Bayesian Optimization of Expensive Cost Functions, with Application to Active User Modeling and Hierarchical Reinforcement Learning , 2010, ArXiv.

[16]  Xiaolu Zhou,et al.  Understanding Spatiotemporal Patterns of Biking Behavior by Analyzing Massive Bike Sharing Data in Chicago , 2015, PloS one.

[17]  Adel W. Sadek,et al.  A k Nearest Neighbor based Local Linear Wavelet Neural Network Model for On-line Short-term Traffic Volume Prediction , 2013 .

[18]  Pan Liu,et al.  The station-free sharing bike demand forecasting with a deep learning approach and large-scale datasets , 2018, Transportation Research Part C: Emerging Technologies.

[19]  Eleni I. Vlahogianni,et al.  Spatio‐Temporal Short‐Term Urban Traffic Volume Forecasting Using Genetically Optimized Modular Networks , 2007, Comput. Aided Civ. Infrastructure Eng..

[20]  Ta-Hui Yang,et al.  A hub location inventory model for bicycle sharing system design: Formulation and solution , 2013, Comput. Ind. Eng..

[21]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[22]  Jieping Ye,et al.  Deep Multi-View Spatial-Temporal Network for Taxi Demand Prediction , 2018, AAAI.

[23]  R. Tibshirani,et al.  Regression shrinkage and selection via the lasso: a retrospective , 2011 .

[24]  Joan Bruna,et al.  Spectral Networks and Locally Connected Networks on Graphs , 2013, ICLR.

[25]  Yu Zheng,et al.  Traffic prediction in a bike-sharing system , 2015, SIGSPATIAL/GIS.

[26]  W. Deng,et al.  Exploring bikesharing travel time and trip chain by gender and day of the week , 2015 .

[27]  Lei Zhu,et al.  Bi-National Delay Pattern Analysis For Commercial and Passenger Vehicles at Niagara Frontier Border , 2017, ArXiv.

[28]  Adel W. Sadek,et al.  Data Mining and Complex Network Algorithms for Traffic Accident Analysis , 2014 .

[29]  Adel W. Sadek,et al.  Short-Term Forecasting of Traffic Volume , 2013 .

[30]  Adel W. Sadek,et al.  On-line prediction of border crossing traffic using an enhanced Spinning Network method , 2014 .

[31]  Yu Zhang,et al.  Free-floating bike sharing: Solving real-life large-scale static rebalancing problems , 2017 .

[32]  Nikhil Ketkar,et al.  Convolutional Neural Networks , 2021, Deep Learning with Python.

[33]  Yunpeng Wang,et al.  Long short-term memory neural network for traffic speed prediction using remote microwave sensor data , 2015 .

[34]  José M. F. Moura,et al.  Discrete Signal Processing on Graphs , 2012, IEEE Transactions on Signal Processing.

[35]  Lei Zhu,et al.  Prediction of Individual Social-Demographic Role Based on Travel Behavior Variability Using Long-Term GPS Data , 2017 .

[36]  Tony Hürlimann,et al.  Modeling Framework , 2021, Hybrid Feedback Control.

[37]  Adel W. Sadek,et al.  Quantifying uncertainty in short-term traffic prediction and its application to optimal staffing plan development , 2018, Transportation Research Part C: Emerging Technologies.

[38]  Yu Zheng,et al.  Deep Spatio-Temporal Residual Networks for Citywide Crowd Flows Prediction , 2016, AAAI.

[39]  Jie Bao,et al.  Exploring Bikesharing Travel Patterns and Trip Purposes Using Smart Card Data and Online Point of Interests , 2017 .

[40]  Pascal Frossard,et al.  The emerging field of signal processing on graphs: Extending high-dimensional data analysis to networks and other irregular domains , 2012, IEEE Signal Processing Magazine.

[41]  Wanli Min,et al.  Real-time road traffic prediction with spatio-temporal correlations , 2011 .

[42]  D. T. Lee,et al.  Travel-time prediction with support vector regression , 2004, IEEE Transactions on Intelligent Transportation Systems.

[43]  Xavier Bresson,et al.  Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering , 2016, NIPS.

[44]  Wei Shen,et al.  Real-time road traffic forecasting using regime-switching space-time models and adaptive LASSO , 2012 .

[45]  Mathieu Bastian,et al.  Gephi: An Open Source Software for Exploring and Manipulating Networks , 2009, ICWSM.

[46]  Adel W. Sadek,et al.  Modeling the Impacts of Inclement Weather on Freeway Traffic Speed , 2015 .

[47]  Lei Lin,et al.  A Comprehensive Framework for Dynamic Bike Rebalancing in a Large Bike Sharing Network , 2018, 1806.02890.

[48]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[49]  Jean-Loup Guillaume,et al.  Fast unfolding of communities in large networks , 2008, 0803.0476.