“…Among of the researches on the discounted DTMDPs, most literature such as [1,2,6,9,14,15,21,22] shared a common feature: the discount factor is a constant, which obviously cannot cover the important case of non-constant interest rates in a bank, and which motivates the studies on the general case of varying discount factors. For example, [8] studied the first passage problem for non-stationary nonlinear discrete-time stochastic control systems with the time-and state-dependent discount factors, and proved the existence of an optimal Markov policy under suitable conditions. Recently, [24] considered the first passage model of discounted DTMDPs in Borel state and action spaces with state-dependent discount factors and unbounded rewards, and showed the existence of the first passage optimal policies.…”