Adjoint-aided inference of Gaussian process driven differential equations

Bayesian inference for the truncated GP can be achieved, in many cases with substantially lower computation than would be required using MCMC methods.