在机器人上应用增强学习（RL）方法通常涉及培训模拟和部署现实世界中的机器人的政策。由于现实世界和模拟器之间的模型不匹配，以这种方式部署的RL代理商倾向于逐渐执行。为了解决这个问题，研究人员制定了强大的政策学习算法，依赖于合成噪声干扰。但是，这些方法在目标环境中不保证性能。我们提出了一种凸起风险最小化算法，以估计模拟器和目标域之间的模型不匹配使用来自两个环境的轨迹数据。我们表明该估计器可以随着模拟器使用，以评估目标域中的RL代理的性能，有效地弥合这两个环境之间的差距。我们还表明，我们的估算器的收敛速度为$ {n { - 1/4}} $，其中$ n $是培训样本的数量。在仿真中，我们展示了我们的方法如何有效地近似和评估GridWorld，Cartpole和Reverser环境的性能。我们还表明，我们的方法能够使用模拟器和远程收集来自现实世界中的机器人的远程收集的数据来估计7 DOF机器人手臂的性能。

非政策评估（OPE）方法是评估高风险领域（例如医疗保健）中的政策的关键工具，在这些领域，直接部署通常是不可行的，不道德的或昂贵的。当期望部署环境发生变化（即数据集偏移）时，对于OPE方法，在此类更改中对策略进行强大的评估非常重要。现有的方法考虑对可以任意改变环境的任何可观察到的任何可观察到的属性的大量转变。这通常会导致对公用事业的高度悲观估计，从而使可能对部署有用的政策无效。在这项工作中，我们通过研究领域知识如何帮助提供对政策公用事业的更现实的估计来解决上述问题。我们利用人类的投入，在环境的哪些方面可能会发生变化，并适应OPE方法仅考虑这些方面的转变。具体而言，我们提出了一个新颖的框架，可靠的OPE（绳索），该框架认为基于用户输入的数据中的协变量子集，并估算了这些变化下最坏情况的效用。然后，我们为OPE开发了对OPE的计算有效算法，这些算法对上述强盗和马尔可夫决策过程的上述变化很强。我们还理论上分析了这些算法的样品复杂性。从医疗领域进行的合成和现实世界数据集进行了广泛的实验表明，我们的方法不仅可以捕获现实的数据集准确地转移，而且还会导致较少的悲观政策评估。

脱机强化学习 - 从一批数据中学习策略 - 是难以努力的：如果没有制造强烈的假设，它很容易构建实体算法失败的校长。在这项工作中，我们考虑了某些现实世界问题的财产，其中离线强化学习应该有效：行动仅对一部分产生有限的行动。我们正规化并介绍此动作影响规律（AIR）财产。我们进一步提出了一种算法，该算法假定和利用AIR属性，并在MDP满足空气时绑定输出策略的子优相。最后，我们展示了我们的算法在定期保留的两个模拟环境中跨越不同的数据收集策略占据了现有的离线强度学习算法。

本文研究了马尔可夫决策过程（MDPS）中用于政策评估的数据收集问题。在政策评估中，我们获得了目标政策，并要求估计它将在正式作为MDP的环境中获得的预期累积奖励。我们通过首先得出了使用奖励分布方差知识的Oracle数据收集策略来开发在树结构MDPS中的最佳数据收集理论。然后，我们介绍了减少的方差采样（射击）算法，即当奖励方差未知并与Oracle策略相比，奖励方差未知并绑定其亚典型性时，它近似于Oracle策略。最后，我们从经验上验证了射手会导致与甲骨文策略相当的均衡误差进行政策评估，并且比仅仅运行目标策略要低得多。

Effectively leveraging large, previously collected datasets in reinforcement learning (RL) is a key challenge for large-scale real-world applications. Offline RL algorithms promise to learn effective policies from previously-collected, static datasets without further interaction. However, in practice, offline RL presents a major challenge, and standard off-policy RL methods can fail due to overestimation of values induced by the distributional shift between the dataset and the learned policy, especially when training on complex and multi-modal data distributions. In this paper, we propose conservative Q-learning (CQL), which aims to address these limitations by learning a conservative Q-function such that the expected value of a policy under this Q-function lower-bounds its true value. We theoretically show that CQL produces a lower bound on the value of the current policy and that it can be incorporated into a policy learning procedure with theoretical improvement guarantees. In practice, CQL augments the standard Bellman error objective with a simple Q-value regularizer which is straightforward to implement on top of existing deep Q-learning and actor-critic implementations. On both discrete and continuous control domains, we show that CQL substantially outperforms existing offline RL methods, often learning policies that attain 2-5 times higher final return, especially when learning from complex and multi-modal data distributions.Preprint. Under review.

Learning policies from fixed offline datasets is a key challenge to scale up reinforcement learning (RL) algorithms towards practical applications. This is often because off-policy RL algorithms suffer from distributional shift, due to mismatch between dataset and the target policy, leading to high variance and over-estimation of value functions. In this work, we propose variance regularization for offline RL algorithms, using stationary distribution corrections. We show that by using Fenchel duality, we can avoid double sampling issues for computing the gradient of the variance regularizer. The proposed algorithm for offline variance regularization (OVAR) can be used to augment any existing offline policy optimization algorithms. We show that the regularizer leads to a lower bound to the offline policy optimization objective, which can help avoid over-estimation errors, and explains the benefits of our approach across a range of continuous control domains when compared to existing state-of-the-art algorithms.

We consider the problem of off-policy evaluation (OPE) in reinforcement learning (RL), where the goal is to estimate the performance of an evaluation policy, $\pi_e$, using a fixed dataset, $\mathcal{D}$, collected by one or more policies that may be different from $\pi_e$. Current OPE algorithms may produce poor OPE estimates under policy distribution shift i.e., when the probability of a particular state-action pair occurring under $\pi_e$ is very different from the probability of that same pair occurring in $\mathcal{D}$ (Voloshin et al. 2021, Fu et al. 2021). In this work, we propose to improve the accuracy of OPE estimators by projecting the high-dimensional state-space into a low-dimensional state-space using concepts from the state abstraction literature. Specifically, we consider marginalized importance sampling (MIS) OPE algorithms which compute state-action distribution correction ratios to produce their OPE estimate. In the original ground state-space, these ratios may have high variance which may lead to high variance OPE. However, we prove that in the lower-dimensional abstract state-space the ratios can have lower variance resulting in lower variance OPE. We then highlight the challenges that arise when estimating the abstract ratios from data, identify sufficient conditions to overcome these issues, and present a minimax optimization problem whose solution yields these abstract ratios. Finally, our empirical evaluation on difficult, high-dimensional state-space OPE tasks shows that the abstract ratios can make MIS OPE estimators achieve lower mean-squared error and more robust to hyperparameter tuning than the ground ratios.

Off-policy reinforcement learning aims to leverage experience collected from prior policies for sample-efficient learning. However, in practice, commonly used off-policy approximate dynamic programming methods based on Q-learning and actor-critic methods are highly sensitive to the data distribution, and can make only limited progress without collecting additional on-policy data. As a step towards more robust off-policy algorithms, we study the setting where the off-policy experience is fixed and there is no further interaction with the environment. We identify bootstrapping error as a key source of instability in current methods. Bootstrapping error is due to bootstrapping from actions that lie outside of the training data distribution, and it accumulates via the Bellman backup operator. We theoretically analyze bootstrapping error, and demonstrate how carefully constraining action selection in the backup can mitigate it. Based on our analysis, we propose a practical algorithm, bootstrapping error accumulation reduction (BEAR). We demonstrate that BEAR is able to learn robustly from different off-policy distributions, including random and suboptimal demonstrations, on a range of continuous control tasks.

Offline reinforcement learning (RL) refers to the problem of learning policies entirely from a large batch of previously collected data. This problem setting offers the promise of utilizing such datasets to acquire policies without any costly or dangerous active exploration. However, it is also challenging, due to the distributional shift between the offline training data and those states visited by the learned policy. Despite significant recent progress, the most successful prior methods are model-free and constrain the policy to the support of data, precluding generalization to unseen states. In this paper, we first observe that an existing model-based RL algorithm already produces significant gains in the offline setting compared to model-free approaches. However, standard model-based RL methods, designed for the online setting, do not provide an explicit mechanism to avoid the offline setting's distributional shift issue. Instead, we propose to modify the existing model-based RL methods by applying them with rewards artificially penalized by the uncertainty of the dynamics. We theoretically show that the algorithm maximizes a lower bound of the policy's return under the true MDP. We also characterize the trade-off between the gain and risk of leaving the support of the batch data. Our algorithm, Model-based Offline Policy Optimization (MOPO), outperforms standard model-based RL algorithms and prior state-of-the-art model-free offline RL algorithms on existing offline RL benchmarks and two challenging continuous control tasks that require generalizing from data collected for a different task. * equal contribution. † equal advising. Orders randomized.34th Conference on Neural Information Processing Systems (NeurIPS 2020),

离线政策评估（OPE）被认为是强化学习（RL）的基本且具有挑战性的问题。本文重点介绍了基于从无限 - 马尔可夫决策过程的框架下从可能不同策略生成的预收集的数据的目标策略的价值估计。由RL最近开发的边际重要性采样方法和因果推理中的协变量平衡思想的动机，我们提出了一个新颖的估计器，具有大约投影的国家行动平衡权重，以进行策略价值估计。我们获得了这些权重的收敛速率，并表明拟议的值估计量在技术条件下是半参数有效的。就渐近学而言，我们的结果比例均以每个轨迹的轨迹数量和决策点的数量进行扩展。因此，当决策点数量分歧时，仍然可以使用有限的受试者实现一致性。此外，我们开发了一个必要且充分的条件，以建立贝尔曼操作员在政策环境中的适当性，这表征了OPE的困难，并且可能具有独立的利益。数值实验证明了我们提出的估计量的有希望的性能。

许多连续的决策问题是使用使用其他一些策略收集的历史数据，需要使用历史数据的高赌注并要求新策略（OPE）。提供无偏估计的最常见的OPE技术之一是基于轨迹的重要性采样（是）。但是，由于轨迹的高方差是估计，最近通过了基于国家行动探索分布（SIS）的重要性采样方法。不幸的是，虽然SIS经常为长视野提供较低的方差估计，但估算状态行动分配比可能是具有挑战性的并且导致偏差估计。在本文中，我们对该偏差差异进行了新的视角，并显示了存在终点是SIS的估计频谱的存在。此外，我们还建立了这些估算器的双重强大和加权版本的频谱。我们提供了经验证据，即该频谱中的估计值可用于在IS和SIS的偏差和方差之间进行折衷，并且可以实现比两者和SIS更低的平均平方误差。

依赖于太多的实验来学习良好的行动，目前的强化学习（RL）算法在现实世界的环境中具有有限的适用性，这可能太昂贵，无法探索探索。我们提出了一种批量RL算法，其中仅使用固定的脱机数据集来学习有效策略，而不是与环境的在线交互。批量RL中的有限数据产生了在培训数据中不充分表示的状态/行动的价值估计中的固有不确定性。当我们的候选政策从生成数据的候选政策发散时，这导致特别严重的外推。我们建议通过两个直接的惩罚来减轻这个问题：减少这种分歧的政策限制和减少过于乐观估计的价值约束。在全面的32个连续动作批量RL基准测试中，我们的方法对最先进的方法进行了比较，无论如何收集离线数据如何。

强化学习算法的实用性由于相对于问题大小的规模差而受到限制，因为学习$ \ epsilon $ -optimal策略的样本复杂性为$ \ tilde {\ omega} \ left（| s | s || a || a || a || a | h^3 / \ eps^2 \ right）$在MDP的最坏情况下，带有状态空间$ S $，ACTION SPACE $ A $和HORIZON $ H $。我们考虑一类显示出低级结构的MDP，其中潜在特征未知。我们认为，价值迭代和低级别矩阵估计的自然组合导致估计误差在地平线上呈指数增长。然后，我们提供了一种新算法以及统计保证，即有效利用了对生成模型的访问，实现了$ \ tilde {o} \ left的样本复杂度（d^5（d^5（| s |+| a |）\），我们有效利用低级结构。对于等级$ d $设置的Mathrm {Poly}（h）/\ EPS^2 \ right）$，相对于$ | s |，| a | $和$ \ eps $的缩放，这是最小值的最佳。与线性和低级别MDP的文献相反，我们不需要已知的功能映射，我们的算法在计算上很简单，并且我们的结果长期存在。我们的结果提供了有关MDP对过渡内核与最佳动作值函数所需的最小低级结构假设的见解。

We study the problem of off-policy value evaluation in reinforcement learning (RL), where one aims to estimate the value of a new policy based on data collected by a different policy. This problem is often a critical step when applying RL to real-world problems. Despite its importance, existing general methods either have uncontrolled bias or suffer high variance. In this work, we extend the doubly robust estimator for bandits to sequential decision-making problems, which gets the best of both worlds: it is guaranteed to be unbiased and can have a much lower variance than the popular importance sampling estimators. We demonstrate the estimator's accuracy in several benchmark problems, and illustrate its use as a subroutine in safe policy improvement. We also provide theoretical results on the inherent hardness of the problem, and show that our estimator can match the lower bound in certain scenarios.

值得信赖的强化学习算法应有能力解决挑战性的现实问题，包括{Robustly}处理不确定性，满足{安全}的限制以避免灾难性的失败，以及在部署过程中{prencepentiming}以避免灾难性的失败}。这项研究旨在概述这些可信赖的强化学习的主要观点，即考虑其在鲁棒性，安全性和概括性上的内在脆弱性。特别是，我们给出严格的表述，对相应的方法进行分类，并讨论每个观点的基准。此外，我们提供了一个前景部分，以刺激有希望的未来方向，并简要讨论考虑人类反馈的外部漏洞。我们希望这项调查可以在统一的框架中将单独的研究汇合在一起，并促进强化学习的可信度。

政策梯度（PG）算法是备受期待的强化学习对现实世界控制任务（例如机器人技术）的最佳候选人之一。但是，每当必须在物理系统上执行学习过程本身或涉及任何形式的人类计算机相互作用时，这些方法的反复试验性质就会提出安全问题。在本文中，我们解决了一种特定的安全公式，其中目标和危险都以标量奖励信号进行编码，并且学习代理被限制为从不恶化其性能，以衡量为预期的奖励总和。通过从随机优化的角度研究仅行为者的政策梯度，我们为广泛的参数政策建立了改进保证，从而将现有结果推广到高斯政策上。这与策略梯度估计器的差异的新型上限一起，使我们能够识别出具有很高概率的单调改进的元参数计划。两个关键的元参数是参数更新的步长和梯度估计的批处理大小。通过对这些元参数的联合自适应选择，我们获得了具有单调改进保证的政策梯度算法。

在钢筋学习中，体验重播存储过去的样本以进一步重用。优先采样是一个有希望的技术，可以更好地利用这些样品。以前的优先级标准包括TD误差，近似和纠正反馈，主要是启发式设计。在这项工作中，我们从遗憾最小化目标开始，并获得最佳的贝尔曼更新优先级探讨策略，可以直接最大化策略的返回。该理论表明，具有较高后视TD误差的数据，应在采样期间具有更高权重的重量来分配更高的Hindsight TD误差，更好的政策和更准确的Q值。因此，最先前的标准只会部分考虑这一战略。我们不仅为以前的标准提供了理论理由，还提出了两种新方法来计算优先级重量，即remern并恢复。 remern学习错误网络，而remert利用状态的时间顺序。这两种方法都以先前的优先考虑的采样算法挑战，包括Mujoco，Atari和Meta-World。

离线政策优化可能会对许多现实世界的决策问题产生重大影响，因为在线学习在许多应用中可能是不可行的。重要性采样及其变体是离线策略评估中一种常用的估计器类型，此类估计器通常不需要关于价值函数或决策过程模型功能类的属性和代表性能力的假设。在本文中，我们确定了一种重要的过度拟合现象，以优化重要性加权收益，在这种情况下，学到的政策可以基本上避免在最初的状态空间的一部分中做出一致的决策。我们提出了一种算法，以避免通过新的每个国家 - 邻居标准化约束过度拟合，并提供对拟议算法的理论理由。我们还显示了以前尝试这种方法的局限性。我们在以医疗风格的模拟器为中测试算法，该模拟器是从真实医院收集的记录数据集和连续的控制任务。这些实验表明，与最先进的批处理学习算法相比，所提出的方法的过度拟合和更好的测试性能。

在本文中，我们提出了一个健壮的模仿学习（IL）框架，该框架在扰动环境动态时改善了IL的稳健性。在单个环境中训练的现有IL框架可能会因环境动力学的扰动而灾难性地失败，因为它无法捕获可以更改潜在环境动态的情况。我们的框架有效地处理了具有不同动态的环境，通过模仿了采样环境动力学中的多个专家，以增强环境动力学的一般变化中的鲁棒性。为了强力模仿多个样本专家，我们将代理商政策与每个样本专家之间的Jensen-Shannon分歧降低了风险。数值结果表明，与常规IL基准相比，我们的算法显着提高了针对动力学扰动的鲁棒性。

While reinforcement learning algorithms provide automated acquisition of optimal policies, practical application of such methods requires a number of design decisions, such as manually designing reward functions that not only define the task, but also provide sufficient shaping to accomplish it. In this paper, we view reinforcement learning as inferring policies that achieve desired outcomes, rather than as a problem of maximizing rewards. To solve this inference problem, we establish a novel variational inference formulation that allows us to derive a well-shaped reward function which can be learned directly from environment interactions. From the corresponding variational objective, we also derive a new probabilistic Bellman backup operator and use it to develop an off-policy algorithm to solve goal-directed tasks. We empirically demonstrate that this method eliminates the need to hand-craft reward functions for a suite of diverse manipulation and locomotion tasks and leads to effective goal-directed behaviors.