1: \begin{abstract}

2: A probabilistic approach of computing geometric rate of convergence

3: of stochastic processes is introduced in this paper. The goal is to

4: quantitatively compute  both upper and lower bounds of the exponential

5: rate of convergence to the invariant probability measure of a stochastic process.

6: By applying the coupling method,

7: % Instead of discretizing the infinitesimal,  

8: we derive an algorithm  %to compute  quantitative lower bound. In addition, because of the existence of an optimal coupling, the upper bound of convergence rate is given by a certain first exit time. 

9: which does not need to discretize the state space. In this way,  our approach works well for many high-dimensional examples. %e.g., the $100$-dimensional coupled FizHugh-Nagumo oscillator. 

10: We apply this algorithm to both random perturbed iterative mappings and stochastic differential equations. %by using different circle maps with distinct mixing properties, 

11: We show that the rate of geometric ergodicity of the random perturbed system can, to some extent, reveal the chaotic properties of the unperturbed deterministic one.

12: %For a stochastic differential equations (SDE), we show that our numerical coupling method is convergent to a certain coupling scheme of the original process. 

13: Various SDE models  including those with degenerate noise or high-dimensional state space are also explored. 

14: %including the stochastic slow-fast, SIR models with degenerate noise, as well as the high-dimensional 

15: 

16: 

17: 

18: 

19: 

20: 

21: %the coupling probability  is convergent as the time-step size approaches to zero. 

22: 

23: \end{abstract}

24: