Frost forecasting model using graph neural networks with spatio-temporal attention - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year :

Frost forecasting model using graph neural networks with spatio-temporal attention

Hernan Lira
  • Function : Author
  • PersonId : 1102200
Luis Martí
Nayat Sanchez-Pi
  • Function : Author
  • PersonId : 1102034

Abstract

Frost forecast is an important issue in climate research because of its economic impact in several industries. In this study, a graph neural network (GNN) with spatio-temporal architecture is proposed to predict minimum temperatures in an experimental site. The model considers spatial and temporal relations and processes multiple time series simultaneously. Performing predictions of 6, 12, and 24 hrs this model outperforms statistical and non-graph deep learning models.
Fichier principal
Vignette du fichier
iclr2021_conference.pdf (1.61 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03259658 , version 1 (15-06-2021)

Identifiers

  • HAL Id : hal-03259658 , version 1

Cite

Hernan Lira, Luis Martí, Nayat Sanchez-Pi. Frost forecasting model using graph neural networks with spatio-temporal attention. AI: Modeling Oceans and Climate Change Workshop at ICLR 2021, Nayat Sanchez-Pi; Luis Martí, May 2021, Santiago, Chile. ⟨hal-03259658⟩

Collections

INRIA INRIA-CHILE
429 View
379 Download

Share

Gmail Facebook Twitter LinkedIn More