A new avenue for Bayesian inference with INLA

Integrated Nested Laplace Approximations (INLA) has been a successful approximate Bayesian inference framework since its proposal by Rue et al. (2009). The increased computational efficiency and accuracy when compared with sampling-based methods for like MCMC methods, are some contributors to success. Ongoing research in the INLA methodology implementation thereof R package R-INLA, ensures continued relevance practitioners improved performance applicability of INLA. era big data recent developments, presents an opportunity reformulate aspects classic formulation, achieve even faster inference, numerical stability scalability. improvement is especially noticeable data-rich models. Various examples models, Cox's proportional hazards model, item-response theory spatial model including prediction, three-dimensional fMRI used illustrate gains tangible manner.