Bayesian mixture modelling for characterising environmental exposures and outcomes

摘要: Environmental exposure and outcomes assessment is a great challenge to scientists. Increasingly more detailed data are becoming available understand the nature complexity of relationships involved. The methodology mixture models provides means understand, quantify describe features relation- ships within complex sets. In this thesis, we focussed on number applied problems characterise environmental outcomes, including: assessing interaction between exposures as risk factors for health outcomes; identifying di®ering across region; estab- lishing patterns in size concentration aerosol particles over time. Mixture model approaches address these developed examined their suitability contexts.