Depends on how small the observational noise is. If the observations are extremely informative then the latent realization would be extremely constrained even without the regularization of the latent process and a centered parameterization would be preferable.
betanalpha
35
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Parameter inference for SDE in epidemiological models | 12 | 1495 | May 13, 2022 | |
| Simulation and inference for a stochastic process with a random time of event in rstan | 16 | 1959 | October 5, 2020 | |
| Estimating Parameters for Chaotic Systems (Lotka-Volterra Redux) | 14 | 1653 | August 23, 2018 | |
| Stan for stochastic systems | 2 | 476 | June 28, 2019 | |
| Zero chain movement/mixing in forced ODE model | 32 | 2750 | June 1, 2021 |