Discounted dynamic optimization and Bregman divergence
定义了布雷格曼连续性，并证明一个函数是折扣动态优化问题的最优策略函数当且仅当它是布雷格曼连续的。
We define the concept of Bregman continuity and show that a function is an optimal policy function of a discounted dynamic optimization problem if and only if it is Bregman continuous.