1: \begin{abstract}

2: We consider the optimization of Markovian dynamics to pursue 

3: the fastest convergence to the stationary state.

4: The brachistochrone method is applied to the continuous-time master equation 

5: for finite-size systems.

6: The principle of least action leads to a brachistochrone equation 

7: for the transition-rate matrix.

8: Three-state systems are explicitly analyzed, and we find that 

9: the solution violates the detailed balance condition.

10: The properties of the solution are studied in detail to observe 

11: the optimality of the solution.

12: We also discuss the counterdiabatic driving for the Markovian dynamics.

13: The transition-rate matrix is then divided into two parts, and 

14: the state is given by an eigenstate of the first part.

15: The second part violates the detailed balance condition and 

16: plays the role of a counterdiabatic term.

17: \end{abstract}

18: