mdps

Parallel Q-learning Algorithms for MDPs Based on Performance Potentials

一种MDP基于性能势的并行Q学习算法

S (λ): A reinforcement learning algorithm based on average-payoff MDPs

S（λ）：一个基于平均奖赏MDPs的激励学习算法

Discounted and Undiscounted MDPs : a Case Study Based on SARSA (λ) Algorithms

折扣与无折扣MDPs：一个基于SARSA（λ）算法的实例分析

We discuss the reinforcement learning-based optimization methods of Markov decision processes ( MDPs ) using the Markov performance potentials .

在Markov性能势基础上讨论了一种基于强化学习的马尔可夫决策过程（MDP）优化方法。

On-Policy Modeless Reinforcement Learning Algorithms for Average - Payoff MDPs

平均奖赏MDP的在策略无模型激励学习算法

The concept of Markov performance potentials , which is introduced by Cao , offers a new framework and approach for the optimization of MDPs .

Markov性能势理论的提出，为MDP的优化提供了一种新的理论框架和途径。

This article based on the ideas and methods of GIS , connecting with features of mining area disaster designed the Mining area Disaster Production System ( MDPS ) .

根据地理信息系统的基本思想，结合矿区灾害的特点设计了矿区灾害防治系统（MDPS）。

Motivated by the need of practical large-scale Markov systems , we considered in this paper the learning optimization problems for Markov decision processes ( MDPs ) .

为适应实际大规模Markov系统的需要，讨论Markov决策过程（MDP）基于仿真的学习优化问题。

Through these solutions we can integrate theory implementation of message transmission based on MDPs , to make it be a safe , stable , maintainable message transmission system .

通过对基于MDPs的消息传递的理论实现进行整合，使之成为安全可靠、性能稳定、可维护的消息传递系统。

The Effect of Synthetic Muramyl Dipeptide ( MDP ) and Its Derivatives ( MDPs ) on Immune Function in Mice

胞壁酰二肽及其衍生物对小鼠免疫功能的影响

Many sequential decision problems , such as flexible manufacturing systems , traffic command systems and queuing systems etc. , can be modeled as Markov decision processes ( MDPs ) .

实际生活中的许多序贯决策问题，如柔性制造系统、交通指挥系统、排队系统等，都可以模型化为Markov决策过程（MDP）。

This paper elaborates on the low learning efficiency in reinforcement learning due to improper generalization and random exploration policy under deterministic MDPS and proposes a hierarchical reinforcement learning algorithm based on system model .

针对强化学习算法的状态值泛化和随机探索策略在确定性MDP系统控制中存在着学习效率低的问题，本文提出基于模型的层次化强化学习算法。

In this paper , a neuro-dynamic programming ( NDP ) method is discussed via an actor-critic algorithm for Markov decision processes ( MDPs ) based on the learning of performance potentials .

研究马尔可大决策过程（MDP）在actor-critic模式下，基于性能势学习的神经元动态规划（NDP）方法。

Moderate deviation principles ( MDPs ) are proved for the occupation time process of a super Brownian motion with immigration , where the immigration is governed by the Lebesgue measure .

本文证明了当底空间维数d≥3时，一类带移民超布朗运动占位时过程的中偏差，其移民由Lebesgue测度控制。

The extension of reinforcement learning to MDPs with large state , action space and high complexity has inevitably encountered the problem of the curse of dimensionality , which results in slow convergence and long training time .

传统的强化学习算法应用到大状态、动作空间和任务复杂的马尔可夫决策过程问题时，存在收敛速度慢，训练时间长等问题。

The problems on discounting reinforcement learning are analyzed . Several experiments have been performed for comparing the influence of different discounting factors on SARSA (λ) algorithm based on MDPs . The role of average reward scalar to undiscounted SARSA (λ) algorithm is also discussed .

分析了折扣激励学习存在的问题，对MDPs的SARSA（λ）算法进行了折扣的比较实验分析，讨论了平均奖赏常量对无折扣SARSA（（）算法的影响。