Convergence rates of posterior distributions for Brownian semimartingale models

F.H. Van Der Meulen, Aad W. Van Der Vaart, J.H. Van Zanten
2006 Bernoulli  
We consider the asymptotic behavior of posterior distributions based on continuous observations from a Brownian semimartingale model. We present a general result that bounds the posterior rate of convergence in terms of the complexity of the model and the amount of prior mass given to balls centered around the true parameter. This result is illustrated for three special cases of the model: the Gaussian white-noise model, the perturbed dynamical system and the ergodic diffusion model. Some examples for specific priors are discussed as well.
doi:10.3150/bj/1161614950 fatcat:2xmtxocb4belthkzrvrkczplsq