[ABE-L] Workshop in Bayesian Methods for Health Research

[Marcos Prates]

Caros colegas,

Gostaria de convidar a todos a participar do Workshop in Bayesian Methods
for Health Research que será realizado no Instituto de Ciência Exatas da
Universidade Federal de Minas Gerais em Belo Horizonte no dia 17 de Janeiro
de 2025 no Auditório da Estatística (sala 2076) das 13:30 às 15:30.

O evento tem a seguinte programação:

13:30 - 14:25 - Jessica Pavani (University of Calgary, Canada): Exploring
temporal dynamics in spatial random partitions driven by spanning trees: an
application to mosquito-borne diseases

Abstract: Spatially constrained clustering is a significant area of
research, especially when addressing changes over time. Dividing a map into
partitions is a complex task due to the huge number of potential partitions
within the search space. This complexity is heightened in spatio-temporal
clustering, where we must account for sequences of partitions. To address
these challenges, we present a Bayesian model designed for time-dependent
sequences of spatial random partitions, introducing a prior distribution
based on product partition models that establish correlations among
partitions. Furthermore, we utilize random spanning trees to help navigate
the partition search space and ensure spatially constrained clustering.
This research is motivated by a relevant practical issue: detecting spatial
and temporal patterns of mosquito-borne diseases. Given the overdispersion
common in such data, we propose a spatio-temporal Poisson mixture model in
which the mean and dispersion parameters are influenced by spatio-temporal
covariates. We apply this model to analyze weekly reported dengue cases
from 2018 to 2023 in the Southeast region of Brazil and evaluate the
performance of the model using simulated data. Overall, our proposed model
has demonstrated its competitiveness in analyzing the temporal dynamics of
spatial clustering.

14:25 - 14:35 - Break

14:35 - 15:30 - Danilo Alvares (University of Cambridge, United Kingdom): A
Bayesian joint model of multiple longitudinal and categorical outcomes with
application to multiple myeloma using permutation-based variable importance

Abstract: Joint models have proven to be an effective approach for
uncovering potentially hidden connections between various types of
outcomes, mainly continuous, time-to-event, and binary. Typically,
longitudinal continuous outcomes are characterized by linear mixed-effects
models, survival outcomes are described by proportional hazards models, and
the link between outcomes are captured by shared random effects. Other
modeling variations include generalized linear mixed-effects models for
longitudinal data and logistic regression when a binary outcome is present,
rather than time until an event of interest. However, in a clinical
research setting, one might be interested in modeling the physician's
chosen treatment based on the patient's medical history in order to
identify prognostic factors. In this situation, there are often multiple
treatment options, requiring the use of a multiclass classification
approach. Inspired by this context, we develop a Bayesian joint model for
longitudinal and categorical data. In particular, our motivation comes from
a multiple myeloma study, in which biomarkers display nonlinear
trajectories that are well captured through bi-exponential submodels, where
patient-level information is shared with the categorical submodel. We also
present a variable importance strategy for ranking prognostic factors. We
apply our proposal and a competing model to the multiple myeloma data,
compare the variable importance and inferential results for both models,
and illustrate patient-level interpretations using our joint model.

Atenciosamente,

--marcos
-------------- Próxima Parte ----------
Um anexo em HTML foi limpo...
URL: <http://lists.ime.usp.br/pipermail/abe/attachments/20250110/5f01b622/attachment.htm>