1: \begin{abstract}

2: Stochastic evolution equations describing the dynamics of systems under the influence of both deterministic and stochastic forces are prevalent in all fields of science.

3: Yet, identifying these systems from sparse-in-time observations remains still a challenging endeavour.

4: Existing approaches focus either on the temporal structure of the observations by relying on conditional expectations, discarding thereby information ingrained in the geometry of the system's invariant density; or employ geometric approximations of the invariant density, which are nevertheless restricted to systems with conservative forces. 

5: Here we propose a method that reconciles these two paradigms. We introduce a new data-driven path augmentation scheme that takes the local observation geometry into account. By employing non-parametric inference on the augmented paths, we can efficiently identify the deterministic driving forces of the underlying system for systems observed at low sampling rates. %Our method highlights the need for introducing notions from the field of geometric statistics for inference of dynamical stochastic models, and pinpoints the need for further study of the (largely unexplored) geometric properties of invariant densities induced by stochastic dynamical systems.

6: 

7:  %We introduce a method that provides fast convergence and is relevant also for dissipative systems. We employ geometry informed conditional expectations by means of a novel data augmentation procedure that .

8: 

9:  %we employ geometrically informed path augmentation to perform Bayesian inference to recover the drift when observations are sparse.

10: 

11: %Constrained to be close to most probable path in the Onsager-Machlup sense for the empirical distribution in the path space

12: \end{abstract}

13: