Skip to main navigation Skip to search Skip to main content

Bayesian Zero-Inflated Joint Modelling of Impacts of Wildlife Trade on Global Emerging Infectious Diseases with INLA Approach

    Activity: Supervision › Master Dissertation Supervision

    Description

    Emerging Infectious Disease (EID) events pose significant threats to public health,
    the global economy, and human activities. Recent EIDs, such as SARS, H1N1 influenza,
    and COVID-19, have resulted in millions of excess deaths worldwide and
    substantial economic losses. Given that the most EIDs are zoonotic, and the unclear
    influence of wildlife trade on the EIDs event with increasing human-animal contact.
    This study investigates the influence of wildlife trade on EIDs after adjusting for the
    effects of socio-economic factors, meteorological factors, geographical factors, and
    biological factors under Bayesian framework. The data is principally available from
    CITES, EIDR and Word Bank during 1975-2013.
    Prior to model analysis, we first adjusted EIDs using reporting efforts across countries.
    Global spatial autocorrelation and temporal autocorrelation of the adjusted
    EIDs are analyzed using Moran’s I and ACF plots, respectively. After that, the
    Bayesian zero-inflation joint model is constructed to analyze the zero-inflated EIDs
    with using the faster INLA algorithm compared to MCMC. Penalized complexity
    prior distribution is employed for the hyper-parameters involved in the random effects
    for the trade-off of model bias and variance. The optimal model based on
    common Bayesian criteria indicates that an increase in import wildlife trade activities
    significantly elevates the likelihood of adjusted EIDs but does not impact its
    intensity given its occurrence. Furthermore, GDP and precipitation exhibit a significant
    negative association, while population density, population growth rate, and
    mammal richness show a significant positive association with adjusted EIDs.
    Period2023