Hi,
I understand you are using the Bayesian state space model and want to know whether you need to put the state matrix with values for discrete or continuous mode. This will depend on the nature of your data and the specific application, as the Bayesian model can have both modes depending on it. If you're working with time series data that are observed at regular discrete intervals, you should use a discrete state space model. If the data or the underlying processes you're modelling are inherently continuous, you might use a continuous model.
You can define a general Bayesian state space model in the following way:
Mdl = bssm(@paramMap,@priorDistribution)
where “paramMap” is parameter-to-matrix mapping function, stored as a function handle and “priorDistribution” is log density of parameter prior distribution.
You might refer to the following pages for more information:
- https://www.mathworks.com/help/econ/bssm.html (Documentation for Bayesian state space models in MATLAB)
- https://www.mathworks.com/matlabcentral/answers/409026-discretizing-a-continuous-state-space-model (MATLAB Answer that talks about conversion between discrete and continuous Bayesian state space models)
I hope this helps!
