Multivariate Time Series Deep Spatiotemporal Forecasting with Graph Neural Network

Multivariate time series forecasting has long been a subject of great concern. For example, there are many valuable applications in electricity consumption, solar power generation, traffic congestion, finance, and so on. Accurately periodic data such as can greatly improve the reliability tasks engineering applications. Time problems often modeled using deep learning methods. However, information sequences dependencies among multiple variables not fully utilized existing Therefore, multivariate spatiotemporal model with graph neural network (MDST-GNN) is proposed to solve shortcomings accuracy prediction this paper. This integrates information. It comprises four modules: learning, temporal convolution, down-sampling convolution. The module extracts between variables. convolution abstracts each variable sequence. used for fusion structure module. An attention mechanism presented filter different sparsities. To verify effectiveness model, experiments carried out on datasets. Experimental results show that outperforms current state-of-the-art baseline solving problem verified by ablation experiments.