Projeto: DSAIPA/CS/0086/2020

Designação do projeto:	MLDLCOV - Impacto das medidas de confinamento relativas ao COVID-19 sobre mobilidade, poluição do ar, saúde e indicadores macroeconómicos em Portugal: uma abordagem em Aprendizado de Máquina
Código do projeto:	DSAIPA/CS/0086/2020
Objetivo Principal:	Reforçar a investigação, o desenvolvimento tecnológico e a inovação
Região de Intervenção:	Norte
Instituição proponente/ Promotor líder/ Entidade coordenadora:	Faculdade de Engenharia da Universidade do Porto
Parceiro(s) / Co-promotor(es) / Instituição(ões) participante(s):	Administração Regional De Saúde Do Norte I.P.; Câmara Municipal da Maia; Cruz Vermelha Portuguesa ; Faculdade de Economia da Universidade do Porto; Instituto de Sistemas e Robotica
Data de aprovação:	2020-12-16
Data de início:	2021-04-01
Data de conclusão:	2024-03-31

Objetivos, atividades e resultados esperados/atingidos

Objectives:

Quantify the impact of different confinement measures implemented in Portugal on the following predefined variables: mobility, air pollution, health, and macroeconomic indicators.
Classify or predict the evolution of the various predefined variables as targets of the developed models, with emphasis on forecasting the evolution of the Portuguese economy.
Analyze the improvement in air quality promoted by COVID-19 related confinement measures.
Understand the effects of the response to the COVID-19 epidemic on air pollution and mortality rates.
Activities:

Collect and integrate data from different sources into a single database.
Spatial analysis of data to understand the spread of the COVID-19 epidemic in Portugal.
Evaluate the impact of confinement measures through Granger causality analysis, advanced regularization techniques based on Machine Learning, and the Augmented Synthetic Control method.
Construct a multivariate panel database by district using data collected from Google, and by municipality using data collected from the National Institute of Statistics (INE), to analyze the effects of a specific local treatment, such as the sanitary cord imposed in Ovar.
Develop models based on recurrent neural networks (RNN) with attention mechanism, convolutional neural networks (CNN) such as WaveNet optimized for studying multivariate time series, and AC-GAN models to simulate scenarios based on sequence-to-sequence generators.
Conduct comparative studies between Deep Learning models and econometric models.
Present studies at conferences and seminars.
Prepare a final report that incorporates the main conclusions of the project.
Share information with stakeholders involved in the project.
Expected Results:

Quantitative assessment of the impact of confinement measures in Portugal on the predefined variables.
Developed models to classify or predict the evolution of the predefined variables.
In-depth analysis of the air quality improvement associated with confinement measures.
Identification of air pollutants that decreased with confinement measures.
Estimation of health benefits resulting from the reduction in air pollution levels.
Contribution to the understanding of the effects of the response to the COVID-19 epidemic on air pollution, economy, and public health.
Publication of scientific articles in international journals.

Recomendar Página Voltar ao Topo

Copyright 1996-2024 © Faculdade de Economia da Universidade do Porto I Termos e Condições I Acessibilidade I Índice A-Z I Livro de Visitas
Página gerada em: 2024-09-28 às 03:38:04 | Política de Utilização Aceitável | Política de Proteção de Dados Pessoais | Denúncias