期刊
ENTROPY
卷 24, 期 11, 页码 -出版社
MDPI
DOI: 10.3390/e24111651
关键词
time-series; spatiotemporal information transformation; attention mechanism; transformer network
资金
- National Science and Technology Major Project
- Natural Science Foundation of China
- Strategic Priority Project of Chinese Academy of Sciences
- Sichuan Science and Technology Program
- [2018ZX10201002]
- [2017YFA0505500]
- [2021YFF1201200]
- [31930022]
- [12131020]
- [12026608]
- [XDB38040400]
- [2022YFS0048]
This study proposes a novel spatiotemporal transformer neural network (STNN) for efficient prediction of high-dimensional short-term time-series, which utilizes a continuous attention mechanism and various attention mechanisms to integrate and predict effective information. Experimental results demonstrate that STNN significantly outperforms existing methods in multi-step forecasting.
Predicting high-dimensional short-term time-series is a difficult task due to the lack of sufficient information and the curse of dimensionality. To overcome these problems, this study proposes a novel spatiotemporal transformer neural network (STNN) for efficient prediction of short-term time-series with three major features. Firstly, the STNN can accurately and robustly predict a high-dimensional short-term time-series in a multi-step-ahead manner by exploiting high-dimensional/spatial information based on the spatiotemporal information (STI) transformation equation. Secondly, the continuous attention mechanism makes the prediction results more accurate than those of previous studies. Thirdly, we developed continuous spatial self-attention, temporal self-attention, and transformation attention mechanisms to create a bridge between effective spatial information and future temporal evolution information. Fourthly, we show that the STNN model can reconstruct the phase space of the dynamical system, which is explored in the time-series prediction. The experimental results demonstrate that the STNN significantly outperforms the existing methods on various benchmarks and real-world systems in the multi-step-ahead prediction of a short-term time-series.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据