A path integral approach to Bayesian inference in Markov processes

We formulate Bayesian updates in Markov processes by means of path integral techniques and derive the imaginary-time Schr"{o}dinger equation with likelihood to direct the inference incorporated as a potential for the posterior probability distribution