Daniel Simpson

Research Statement

AI generated · Oct 26, 2026

Daniel Simpson's indexed work centers on making Bayesian spatial statistics computationally tractable. His most-cited first-author paper, penalising model component complexity to construct principled, practical priors (2017), has drawn over a thousand citations, while computationally efficient inference for log-Gaussian Cox processes off the grid (2016) and forgetting the covariance function to make spatial statistics feasible (2011) pursue the same goal of scalable spatial modelling. A related senior-author line, Russian roulette estimates for Bayesian inference with doubly-intractable likelihoods (2015), extends this computational agenda to otherwise intractable models. Across these works, the recurring thread is replacing expensive covariance-based machinery with approximations that keep inference practical. Simpson also contributed as a middle author to widely used Bayesian computing references, including a review of Bayesian computing with INLA (2017) and new features for INLA (2013), alongside an intuitive Bayesian spatial model for disease mapping that accounts for scaling (2016). His broader output spans spatio-temporal modelling of particulate matter via the SPDE approach (2012) and constructing priors that penalize the complexity of Gaussian random fields (2018). No software, datasets, or grants are indexed beyond these publications. Taken together, the named works trace a consistent arc: building the computational and prior-construction foundations that let complex spatial and Bayesian models be fitted in practice.

Most Cited Works

Will my boomerang come back? New insights into Aboriginal material culture of early Sydney and affiliated coastal zone from British collections

Gaye Sculthorpe, Daniel Simpson

Australian Archaeology2023

3Citations

Navigating the Customs House, Then and Now: A Synthesis of British Colonial Collecting in Australia, 1788–1823

Daniel Simpson

Australian Historical Studies2023

1Citations

Research Areas

Gaussian Processes and Bayesian InferencMarkov Chains and Monte Carlo MethodsSpatial and Panel Data AnalysisStatistical Methods and Bayesian InferenPoint processes and geometric inequaliti

Loading ORCID activity...

Sources

Research Statement

Most Cited Works

Will my boomerang come back? New insights into Aboriginal material culture of early Sydney and affiliated coastal zone from British collections

Navigating the Customs House, Then and Now: A Synthesis of British Colonial Collecting in Australia, 1788–1823

Research Areas