Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ Universidade do Minh...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
addClaim

Aplicações de modelos lineares generalizados mistos no estudo de acidentes rodoviários

Authors: Maia, Ana Maria Lavandeira Ferreira;

Aplicações de modelos lineares generalizados mistos no estudo de acidentes rodoviários

Abstract

Os Modelos Lineares Generalizados Mistos (MLGM) são uma extensão dos Modelos Lineares Generalizados (MLG) que permitem incluir componentes de variabilidade não explicada pelos efeitos fixos, sendo adequados para modelar dados de contagem e analisar correlações em dados longitudinais. Estes modelos incorporam efeitos aleatórios no preditor linear, permitindo modelar a estrutura de correlação entre observações do mesmo indivíduo, o que é especialmente relevante em dados longitudinais. Neste trabalho, demonstrou-se a flexibilidade destes modelos na análise de dados de acidentes rodoviários sendo, aplicandos para compreender a influência de diversas variáveis na frequência de ocorrências de acidentes. Os dados são referentes a sublanços das autoestradas A4 e A41 na Região do Grande Porto, Portugal, abrangendo um período de oito anos (2014-2021) e disponibilizados pela concessionária Ascendi. A análise exploratória destacou a A41 como a autoestrada com maior volume de tráfego e maior do número de acidentes, nomeadamente, no sentido crescente. Em 2020, observou-se uma redução acentuada do número de acidentes devido à pandemia e ao confinamento. A irregularidade do pavimento aumentou entre 2014 e 2018, enquanto que a aderência à superfície melhorou. O sublanço ”Maia-Alfena” foi identificado como o sublanço de maior frequência de acidentes, em particular no sentido crescente. A construção de modelos de regressão de Poisson com dois efeitos aleatórios permitiu capturar a heterogeneidade dos dados, revelando que o estado do pavimento, o número de vias e as características geométricas , tanto em planta como em perfil da estrada, influenciam a ocorrência de acidentes.

Generalized Linear Mixed Models (GLMs) are an extension of Generalized Linear Models (GLMs) that allow the inclusion of components of variability not explained by fixed effects, and are suitable for modeling count data and analyzing correlations in longitudinal data. These models incorporate random effects in the linear predictor, allowing the modeling of the correlation structure between observations of the same individual, which is especially relevant in longitudinal data. In this work, the flexibility of these models in the analysis of road accident data was demonstrated, being applied to understand the influence of several variables on the frequency of accident occurrences. The data refer to subsections of the A4 and A41 motorways in the Greater Porto Region, Portugal, covering an eight-year period (2014-2021) and provided by the concessionaire Ascendi. The exploratory analysis highlighted the A41 as the motorway with the highest volume of traffic and the highest number of accidents, particularly in the upward direction. In 2020, there was a sharp reduction in the number of accidents due to the pandemic and lockdown. Pavement unevenness increased between 2014 and 2018, while surface adhesion improved. The ”Maia-Alfena”subsection was identified as the subsection with the highest frequency of accidents, particularly in the ascending direction. The Poisson Regression models with two random effects allowed capturing the heterogeneity of the data, revealing that the condition of the pavement, the number of lanes and the geometric characteristics, both in plan and in the profile of the road, influence the occurrence of accidents.

Country
Portugal
Related Organizations
Keywords

Modelos Lineares Generalizados Mistos, Poisson Regression, Modelação estatística, Road accidents, Acidentes rodoviários, Segurança Rodoviária, Road Safety, Regressão de Poisson, Generalized Linear Mixed Models, Statistical modeling, Ciências Naturais::Matemáticas

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green