1: \begin{abstract}%

2: %%%%%%%%%%%%%%%%%

3: In this paper a class of dynamical systems 

4: describing expectation variables 

5: exactly derived from 

6: continuous-time master equations is introduced and studied from 

7: the viewpoint of differential geometry, where 

8: such master equations consist of a set of appropriately chosen 

9: Markov kernels.

10: To geometrize such dynamical systems for expectation variables, 

11: information geometry is used for expressing equilibrium states, 

12: and contact geometry is used for nonequilibrium states. 

13: Here time-developments of the expectation variables 

14: are identified with contact Hamiltonian vector fields on a contact manifold.

15: Also, it is shown that the convergence rate of this dynamical system 

16: is exponential.  Duality emphasized in information geometry is also 

17: addressed throughout. 

18: %%%%%%%%%%%%%%% 

19: \end{abstract}%