A deep learning approach for multi-attribute data: A study of train delay prediction in railway systems

Ping Huang, Chao Wen, Liping Fu, Qiyuan Peng, Yixiong Tang

Abstract: Highlights•A hybrid deep learning approach was proposed for multi-attribute data.•Data with different formats were processed by specific neural units.•Temporal and spatiotemporal relations were captured by the proposed model.•The model had small errors in train delay prediction.•The model was robust for modeling complex data with different sizes and dimensions.