Making nonlinear non-Gaussian state space models accessible to Masters level statisticians

Ed Ionides

A 5-minute summary of three courses

1. A one-semester time series course for Masters level statisticians:
   • First half of the course: a likelihood-based approach to linear time series models.
   • Second half: likelihood-based investigation of nonlinear state space models, using plug-and-play methodology in the R package pomp.
   
   
2. A three-day (15 classroom hours) course for graduate students interested in state space modeling in epidemiology.
   
   
3. A six-lecture (8 classroom hours) course on nonlinear non-Gaussian state space models for PhD students.