BIROn - Birkbeck Institutional Research Online

    Bayesian nonparametric models for spatially indexed data of mixed type

    Papageorgiou, Georgios and Richardson, S. and Best, N. (2015) Bayesian nonparametric models for spatially indexed data of mixed type. Journal of the Royal Statistical Society - Series B (Statistical Methodology) 77 (5), pp. 973-999. ISSN 1369-7412.

    [img]
    Preview
    Text
    BNPSM.pdf - Author's Accepted Manuscript

    Download (1MB) | Preview

    Abstract

    We develop Bayesian nonparametric models for spatially indexed data of mixed type. Our work is motivated by challenges that occur in environmental epidemiology, where the usual presence of several confounding variables that exhibit complex interactions and high correlations makes it difficult to estimate and under- stand the effects of risk factors on health outcomes of interest. The modeling approach we adopt assumes that responses and confounding variables are manifestations of continuous latent variables, and uses mul- tivariate Gaussians to jointly model these. Responses and confounding variables are not treated equally as relevant parameters of the distributions of the responses only are modeled in terms of explanatory variables or risk factors. Spatial dependence is introduced by allowing the weights of the nonparametric process priors to be location specific, obtained as probit transformations of Gaussian Markov random fields. Con- founding variables and spatial configuration have a similar role in the model, in that they only in uence, along with the responses, the allocation probabilities of the areas into the mixture components, thereby allowing for exible adjustment of the effects of observed confounders, while allowing for the possibility of residual spatial structure, possibly occurring due to unmeasured or undiscovered spatially varying factors. Aspects of the model are illustrated in simulation studies and an application to a real data set.

    Metadata

    Item Type: Article
    Keyword(s) / Subject(s): Latent variables, Multiple confounders, Multiple responses, Probit stick-breaking process, Spatial dependence
    School: Birkbeck Faculties and Schools > Faculty of Science > School of Computing and Mathematical Sciences
    Depositing User: Georgios Papageorgiou
    Date Deposited: 06 Jan 2015 10:30
    Last Modified: 09 Aug 2023 12:35
    URI: https://eprints.bbk.ac.uk/id/eprint/10720

    Statistics

    Activity Overview
    6 month trend
    306Downloads
    6 month trend
    306Hits

    Additional statistics are available via IRStats2.

    Archive Staff Only (login required)

    Edit/View Item Edit/View Item