Discrete Dynamic Programming with Sensitive Discount Optimality Criteria
نویسندگان
چکیده
منابع مشابه
Sensitive Discount Optimality: Unifying Discounted and Average Reward Reinforcement Learning
Research in reinforcement learning (RL) has thus far concentrated on two optimality criteria: the discounted framework, which has been very well-studied, and the average-reward framework, in which interest is rapidly increasing. In this paper, we present a framework called sensitive discount optimality which ooers an elegant way of linking these two paradigms. Although sensitive discount optima...
متن کاملBiasing Approximate Dynamic Programming with a Lower Discount Factor
Most algorithms for solving Markov decision processes rely on a discount factor, which ensures their convergence. It is generally assumed that using an artificially low discount factor will improve the convergence rate, while sacrificing the solution quality. We however demonstrate that using an artificially low discount factor may significantly improve the solution quality, when used in approx...
متن کاملDiscrete Dynamic Programming
We consider a decision problem taking place over time. In each time period, the single player can take an action by α ∈ A , an action space. All information relevant to the future is incorporated in a state variable y Y ∈ , the state space. The dynamics of y are determined by a transition probability π α ( ’ , ) y y . We define the set of states reachable with some probability under some circum...
متن کاملTwo Characterizations of Optimality in Dynamic Programming
It holds in great generality that a plan is optimal for a dynamic programming problem, if and only if it is “thrifty” and “equalizing.” An alternative characterization of an optimal plan, that applies in many economic models, is that the plan must satisfy an appropriate Euler equation and a transversality condition. Here we explore the connections between these two characterizations.
متن کاملOptimality criteria for deterministic discrete-time infinite horizon optimization
We consider the problem of selecting an optimality criterion, when total costs diverge, in deterministic infinite horizon optimization over discrete time. Our formulation allows for both discrete and continuous state and action spaces, as well as time-varying, that is, nonstationary, data. The task is to choose a criterion that is neither too overselective, so that no policy is optimal, nor too...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: The Annals of Mathematical Statistics
سال: 1969
ISSN: 0003-4851
DOI: 10.1214/aoms/1177697379