GTrans: Spatiotemporal Autoregressive Transformer with Graph Embeddings for Nowcasting Extreme Events

01/18/2022
by   Bo Feng, et al.
6

Spatiotemporal time series nowcasting should preserve temporal and spatial dynamics in the sense that generated new sequences from models respect the covariance relationship from history. Conventional feature extractors are built with deep convolutional neural networks (CNN). However, CNN models have limits to image-like applications where data can be formed with high-dimensional arrays. In contrast, applications in social networks, road traffic, physics, and chemical property prediction where data features can be organized with nodes and edges of graphs. Transformer architecture is an emerging method for predictive models, bringing high accuracy and efficiency due to attention mechanism design. This paper proposes a spatiotemporal model, namely GTrans, that transforms data features into graph embeddings and predicts temporal dynamics with a transformer model. According to our experiments, we demonstrate that GTrans can model spatial and temporal dynamics and nowcasts extreme events for datasets. Furthermore, in all the experiments, GTrans can achieve the highest F1 and F2 scores in binary-class prediction tests than the baseline models.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
04/12/2021

Learning dynamic and hierarchical traffic spatiotemporal features with Transformer

Traffic forecasting is an indispensable part of Intelligent transportati...
research
02/10/2021

NAST: Non-Autoregressive Spatial-Temporal Transformer for Time Series Forecasting

Although Transformer has made breakthrough success in widespread domains...
research
10/24/2018

Multistep Speed Prediction on Traffic Networks: A Graph Convolutional Sequence-to-Sequence Learning Approach with Attention Mechanism

Multistep traffic forecasting on road networks is a crucial task in succ...
research
09/14/2022

Scalable Spatiotemporal Graph Neural Networks

Neural forecasting of spatiotemporal time series drives both research an...
research
02/27/2022

Short-term passenger flow prediction for multi-traffic modes: A residual network and Transformer based multi-task learning method

With the prevailing of mobility as a service (MaaS), it becomes increasi...
research
03/21/2023

DG-Trans: Dual-level Graph Transformer for Spatiotemporal Incident Impact Prediction on Traffic Networks

The prompt estimation of traffic incident impacts can guide commuters in...
research
07/03/2021

Spatiotemporal convolutional network for time-series prediction and causal inference

Making predictions in a robust way is not easy for nonlinear systems. In...

Please sign up or login with your details

Forgot password? Click here to reset