a multivariate time series deep spatiotemporal forecasting model with graph neural network (MDST-GNN) is proposed to solve the existing shortcomings and improve the accuracy of periodic data prediction in this paper.

This is a PyTorch implementation of the paper: Multivariate Time Series Deep Spatiotemporal Forecasting with Graph Neural Network, published in Applied Sciences.

Figure 1 Model structure diagram of MDST-GNN.

The multivariate time series dataset Solar-Energy, Traffic, Electricity, Exchange-rate datasets can be found in the data file.

The model is implemented using Python with dependencies specified in requirements.txt