Efficient Gaussian process-based inference for modelling spatio-temporal dengue fever

Julio Albinati, Wagner Meira, Gisele L. Pappa, Andrew G. Wilson

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Dengue fever is a disease that affects hundreds of millions of people every year worldwide. Despite its wide presence around the world, it still requires accurate early warning systems. In this paper, we propose an accurate model to forecast dengue fever incidence at hundreds of Brazilian cities simultaneously. In order to assure efficiency, we devise two strategies to reduce computational effort required for inference under the proposed model. As a result, we not only reduce the computational effort that would be required to fit each model per city, but also increase the accuracy by inducing spatial dependences between cities. These dependences do not require human specification and are learned from data, leading to more accurate predictions than using typical neighborhood or distance-based methods.

Original languageEnglish (US)
Title of host publicationProceedings - 2017 Brazilian Conference on Intelligent Systems, BRACIS 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages61-66
Number of pages6
ISBN (Electronic)9781538624074
DOIs
StatePublished - Jun 28 2017
Event6th Brazilian Conference on Intelligent Systems, BRACIS 2017 - Uberlandia, Brazil
Duration: Oct 2 2017Oct 5 2017

Publication series

NameProceedings - 2017 Brazilian Conference on Intelligent Systems, BRACIS 2017
Volume2018-January

Conference

Conference6th Brazilian Conference on Intelligent Systems, BRACIS 2017
Country/TerritoryBrazil
CityUberlandia
Period10/2/1710/5/17

Keywords

  • Gaussian process
  • dengue fever
  • spatio-temporal modelling

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Hardware and Architecture
  • Control and Optimization

Fingerprint

Dive into the research topics of 'Efficient Gaussian process-based inference for modelling spatio-temporal dengue fever'. Together they form a unique fingerprint.

Cite this