Geilhufe Marc, Held Leonhard, Skrøvseth Stein Olav, Simonsen Gunnar S., Godtliebsen Fred (2014), Power law approximations of movement network data for modeling infectious disease spread, in Biometrical Journal
, 56, 363-382.
Meyer Sebastian, Held Leonhard (2014), Flexible estimation of spatio-temporal interaction in a point process model for infectious disease spread
, Statistical Modelling Society, Göttingen, Germany.
Meyer Sebastian, Held Leonhard (2014), Power-law models for infectious disease spread, in Annals of Applied Statistics
, 8(3), 1612-1639.
Wei Wei, Held Leonhard (2014), Calibration tests for count data, in TEST
, 23(4), 787-805.
Wei Wei, Schüpbach Gertraud, Held Leonhard (2015), Time-series analysis of Campylobacter incidence in Switzerland, in Epidemiology and Infection
, 143, 1982-1989.
Meyer Sebastian, Warnke Ingeborg, Rössler Wulf, Held Leonhard (2016), Model-based testing for space-time interaction using point processes: An application to psychiatric hospital admissions in an urban area, in Spatial and Spatio-temporal Epidemiology
, 17, 15-25.
Meyer Sebastian, Held Leonhard, Höhle Michael (accepted), Spatio-Temporal Analysis of Epidemic Phenomena Using the R Package surveillance, in Journal of Statistical Software
Wei Wei, Balabdaoui Fadoua, Held Leonhard (accepted), Calibration tests for multivariate Gaussian distributions, in Journal of Multivariate Analysis
Summary of proposed research
This research proposal aims to develop novel statistical methodology for both retro- and prospective analysis of space-time data on infectious disease incidence. The new techniques will be applied in the particular context of space-time surveillance data, but important parts of the methodology can be used in a wider context. The proposal is composed of three distinct projects, each of which will contribute significantly to the overall goals of this proposal. The first project is concerned with
appropriate integration of network data in statistical models to better describe the spatio-temporal spread of infectious diseases. Both parameter- and observation-driven models will be suitably extended and statistical algorithms will be proposed to incorporate the impact of network data in
the analysis. Spatially-correlated overdispersion will be allowed for to achieve improved predictions of future disease incidence. The second project is concerned with the prospective analysis of space-time count data. Statistical algorithms for sequential analysis of time series models will be
extended to the space-time setting with particular focus on multivariate outbreak detection. Validation of both retro- and prospective analysis will be based on the assessment of out-of-sample predictions based on proper scoring rules. The third project will develop general methodology to
facilitate this goal. In particular, new techniques for the assessment of multivariate predictions will be developed, which have wider relevance in other application areas.