English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the zaia usage as premium feature

ZAIA

Presents the pdf analysis as premium feature

PDF Analysis

Insights

Presents the summarisation as premium feature

Summarisation

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the access of premium content as premium feature

Premium Content

Zendy Tools

Zendy Open

This study proposes a novel Graph Convolutional Neural Network withData-driven Graph Filter (GCNN-DDGF) model that can learn hidden heterogeneouspairwise correlations between stations to predict station-level hourly demandin a large-scale bike-sharing network. Two architectures of the GCNN-DDGF modelare explored; GCNNreg-DDGF is a regular GCNN-DDGF model which contains theconvolution and feedforward blocks, and GCNNrec-DDGF additionally contains arecurrent block from the Long Short-term Memory neural network architecture tocapture temporal dependencies in the bike-sharing demand series. Furthermore,four types of GCNN models are proposed whose adjacency matrices are based onvarious bike-sharing system data, including Spatial Distance matrix (SD),Demand matrix (DE), Average Trip Duration matrix (ATD), and Demand Correlationmatrix (DC). These six types of GCNN models and seven other benchmark modelsare built and compared on a Citi Bike dataset from New York City which includes272 stations and over 28 million transactions from 2013 to 2016. Results showthat 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 bythe GCNNreg-DDGF. They outperform the other models. Through a more detailedgraph network analysis based on the learned DDGF, insights are obtained on theblack box of the GCNN-DDGF model. It is found to capture some informationsimilar to details embedded in the SD, DE and DC matrices. More importantly, italso uncovers hidden heterogeneous pairwise correlations between stations thatare not revealed by any of those matrices.

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