Notebooks

Learning and capacity building

• naomi: following naomi vignettes
• maths: mathematical description of the Naomi model
• aghq: following aghq vignettes to explore the package
• walkthrough: stepping through aghq code line-by-line
• TMB: note about what TMB is doing
• inla-replication: replicating the “INLA from scratch” section of Spatial and Spatio-temporal Bayesian Models with R-INLA using R and TMB
• inla-grid: illustration of how the hyperparameter posterior marginal is explored in the INLA method
• epil: comparison of Stan, R-INLA, TMB, glmmTMB, tmbstan and aghq for the epilepsy example from Rue, Martino and Chopin (2009)
• prev-anc-art: comparison of tmbstan, TMB and aghq for a collection of models based on Joint small-area estimation of HIV prevalence, ART coverage, and HIV incidence (Eaton et al. 2019)

Grid scale-up

• scale-grid: how can we scale up the number of points in the hyperparameter grid in an intelligent way?
• astro: application of scalable grid to astronomy example from Bilodeau, Stringer and Tang (2022)
• increase-s-k: how does increasing the number of principal components, or number of points per dimension, change estimation of the log normalising constant?

Laplace marginals

• sinla: implementing approximations for the posterior marginal of the latent field, building to the approach of Wood (2021)

Posterior comparison approaches

• posterior-comparison: exploring methods (Kolmogorov-Smirnov, simulation-based calibration, Pareto smoothed importance sampling, maximum mean discrepancy) for comparison of posterior distributions from approximate Bayesian inference methods

Results

• point-estimates: comparison of inference methods for the simplified Naomi model using point estimates (the mean and standard deviation)
• ks: comparison of inference methods for the simplified Naomi model using histograms and KS test results
• exceedance: case-study of exceedance probabilities: probability to meet the second 90 target, high incidence strata, and amount of unmet treatment need
• psis: comparison of inference methods for the simplified Naomi model using Pareto-smoothed importance sampling
• mmd: comparison of inference methods for the simplified Naomi model using maximum mean discrepancy
• mcmc-convergence: assessing MCMC (NUTS using tmbstan) convergence for the simplified Naomi model
• model-checks: checking the fit of the simplified Naomi model to data

Presentations

• 01-04-20-mini: three month CDT mini-project
• 01-07-21-stats-epi-group: Machine Learning & Global Health Network group meeting
• 15-11-22-seminar: Waterloo SAS Student Seminar Series, and abstract
• 18-04-23-lab-group: HIV inference group lab meeting
• 19-05-23-turing: Alan Turing Insitute PhD Student Presentation and Networking
• 21-06-23-mlgh: Machine Learning & Global Health Network group meeting

Posters

• bayescomp-poster: for BayesComp 2023
• bioinference-poster: for BioInference 2023
• statml-poster: for StatML retreat 2023

Manuscript

• paper: work-in-progress write-up
• appendix: additional material for write-up