Disaggregation regression is a statistical method that is useful when you have aggregated disease case data (cases counted over a county for example) but wish to make higher resolution predictions of disease risk to guide interventions such as vaccine deployment or vector control. Current implementations of the model are tailored towards simple linear effects, and in order to make better predictions we wish to develop these methods further to use nonlinear effects. In order to do this, we will be incorporating ideas from MaxEnt species distribution models into disaggregation regression.

These models have many applications, but the focus in this project is on zoonotic and vector-borne diseases with a particular case study of lassa fever. The work will be conducted with guidance from Nigeria CDC in order to ensure that we are building a tool that is useful for the analysts both at Nigeria CDC and at other governmental ministries and NGOs in countries with high burdens of zoonotic or vector-borne diseases.