Bridging the Gap Between Training and Inference for Spatio-Temporal Forecasting

  title={Bridging the Gap Between Training and Inference for Spatio-Temporal Forecasting},
  author={Hongbin Liu and Ickjai Lee},
  • Hongbin Liu, Ickjai Lee
  • Published in ECAI 2020
  • Computer Science, Mathematics
  • Spatio-temporal sequence forecasting is one of the fundamental tasks in spatio-temporal data mining. It facilitates many real world applications such as precipitation nowcasting, citywide crowd flow prediction and air pollution forecasting. Recently, a few Seq2Seq based approaches have been proposed, but one of the drawbacks of Seq2Seq models is that, small errors can accumulate quickly along the generated sequence at the inference stage due to the different distributions of training and… CONTINUE READING
    1 Citations

    Figures, Tables, and Topics from this paper

    MPL-GAN: Toward Realistic Meteorological Predictive Learning Using Conditional GAN
    • PDF


    Deep Uncertainty Learning: A Machine Learning Approach for Weather Forecasting
    • 2
    Deep Forecast: Deep Learning-based Spatio-Temporal Forecasting
    • 48
    • PDF
    Predicting Citywide Crowd Flows Using Deep Spatio-Temporal Residual Networks
    • 126
    • PDF
    Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting
    • 2,535
    • PDF
    Machine Learning for Spatiotemporal Sequence Forecasting: A Survey
    • 24
    • PDF
    Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting
    • 477
    • Highly Influential
    • PDF
    GeoMAN: Multi-level Attention Networks for Geo-sensory Time Series Prediction
    • 103
    • PDF
    Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model
    • 143
    • PDF
    Exploiting Spatio-Temporal Correlations with Multiple 3D Convolutional Neural Networks for Citywide Vehicle Flow Prediction
    • Cen Chen, K. Li, +6 authors Z. Zeng
    • Computer Science
    • 2018 IEEE International Conference on Data Mining (ICDM)
    • 2018
    • 27
    • PDF
    PredRNN++: Towards A Resolution of the Deep-in-Time Dilemma in Spatiotemporal Predictive Learning
    • 88
    • Highly Influential
    • PDF