Approximate inference in latent Gaussian-Markov models from continuous time observations

We propose an approximate inference algorithm for continuous time Gaussian-Markov process models with both discrete and continuous time likelihoods. We show that the continuous time limit of the expec