Many real-world sequential decision-making problems involve critical systems with financial risks and human-life risks. While several works in the past have proposed methods that are safe for deployment, they assume that the underlying problem is stationary. However, many real-world problems of interest exhibit non-stationarity, and when stakes are high, the cost associated with a false stationarity assumption may be unacceptable. We take the first steps towards ensuring safety, with high confidence, for smoothly-varying non-stationary decision problems. Our proposed method extends a type of safe algorithm, called a Seldonian algorithm, through a synthesis of model-free reinforcement learning with time-series analysis. Safety is ensured using sequential hypothesis testing of a policy's forecasted performance, and confidence intervals are obtained using wild bootstrap.
Department of Computer Science, University of Manchester（曼彻斯特大学计算机科学系）；Faculty of Computer Science & DSME, RWTH Aachen University（亚琛工业大学计算机科学与DSME学院）；Department of Computer Science, University of Oxford（牛津大学计算机科学系）；Department of Computing, Imperial College London（伦敦帝国理工学院计算系）
In this paper, we propose new problem-independent lower bounds on the sample complexity and regret in episodic MDPs, with a particular focus on the non-stationary case in which the transition kernel is allowed to change in each stage of the episode. Our main contribution is a novel lower bound of $Ω((H^3SA/ε^2)\log(1/δ))$ on the sample complexity of an $(\varepsilon,δ)$-PAC algorithm for best policy identification in a non-stationary MDP. This lower bound relies on a construction of "hard MDPs" which is different from the ones previously used in the literature. Using this same class of MDPs, w…
A Markov Decision Process (MDP) is a popular model for reinforcement learning. However, its commonly used assumption of stationary dynamics and rewards is too stringent and fails to hold in adversarial, nonstationary, or multi-agent problems. We study an episodic setting where the parameters of an MDP can differ across episodes. We learn a reliable policy of this potentially adversarial MDP by developing an Adversarial Reinforcement Learning (ARL) algorithm that reduces our MDP to a sequence of \emph{adversarial} bandit problems. ARL achieves $O(\sqrt{SATH^3})$ regret, which is optimal with re…
This work focuses on off-policy evaluation (OPE) with function approximation in infinite-horizon undiscounted Markov decision processes (MDPs). For MDPs that are ergodic and linear (i.e. where rewards and dynamics are linear in some known features), we provide the first finite-sample OPE error bound, extending existing results beyond the episodic and discounted cases. In a more general setting, when the feature dynamics are approximately linear and for arbitrary rewards, we propose a new approach for estimating stationary distributions with function approximation. We formulate this problem as …
McGill University（麦吉尔大学）；Mila；University of Massachusetts（马萨诸塞大学）；Facebook AI Research（Facebook人工智能研究院）；Microsoft Research（微软研究院）
We study the problem of Safe Policy Improvement (SPI) under constraints in the offline Reinforcement Learning (RL) setting. We consider the scenario where: (i) we have a dataset collected under a known baseline policy, (ii) multiple reward signals are received from the environment inducing as many objectives to optimize. We present an SPI formulation for this RL setting that takes into account the preferences of the algorithm's user for handling the trade-offs for different reward signals while ensuring that the new policy performs at least as well as the baseline policy along each individual …
We consider a Reinforcement Learning setup where an agent interacts with an environment in observation-reward-action cycles without any (esp.\ MDP) assumptions on the environment. State aggregation and more generally feature reinforcement learning is concerned with mapping histories/raw-states to reduced/aggregated states. The idea behind both is that the resulting reduced process (approximately) forms a small stationary finite-state MDP, which can then be efficiently solved or learnt. We considerably generalize existing aggregation results by showing that even if the reduced process is not an…
Mitsubishi Electric Research Laboratories (MERL)（三菱电机研究实验室）
Safety and robustness are two desired properties for any reinforcement learning algorithm. CMDPs can handle additional safety constraints and RMDPs can perform well under model uncertainties. In this paper, we propose to unite these two frameworks resulting in robust constrained MDPs (RCMDPs). The motivation is to develop a framework that can satisfy safety constraints while also simultaneously offer robustness to model uncertainties. We develop the RCMDP objective, derive gradient update formula to optimize this objective and then propose policy gradient based algorithms. We also independentl…
University of Massachusetts（马萨诸塞大学）；Adobe Research（Adobe研究）；University of Alberta（阿尔伯塔大学）
Most reinforcement learning methods are based upon the key assumption that the transition dynamics and reward functions are fixed, that is, the underlying Markov decision process is stationary. However, in many real-world applications, this assumption is violated, and using existing algorithms may result in a performance lag. To proactively search for a good future policy, we present a policy gradient algorithm that maximizes a forecast of future performance. This forecast is obtained by fitting a curve to the counter-factual estimates of policy performance over time, without explicitly modeli…
RWTH Aachen University（亚琛工业大学）；UMONS – Université de Mons（蒙斯大学）；Vrije Universiteit Brussel（布鲁塞尔自由大学）
We consider the verification of multiple expected reward objectives at once on Markov decision processes (MDPs). This enables a trade-off analysis among multiple objectives by obtaining the Pareto front. We focus on strategies that are easy to employ and implement. That is, strategies that are pure (no randomization) and have bounded memory. We show that checking whether a point is achievable by a pure stationary strategy is NP-complete, even for two objectives, and we provide an MILP encoding to solve the corresponding problem. The bounded memory case can be reduced to the stationary one by a…
Constrained Markov decision processes (CMDPs) model scenarios of sequential decision making with multiple objectives that are increasingly important in many applications. However, the model is often unknown and must be learned online while still ensuring the constraint is met, or at least the violation is bounded with time. Some recent papers have made progress on this very challenging problem but either need unsatisfactory assumptions such as knowledge of a safe policy, or have high cumulative regret. We propose the Safe PSRL (posterior sampling-based RL) algorithm that does not need such ass…
Many physical systems have underlying safety considerations that require that the policy employed ensures the satisfaction of a set of constraints. The analytical formulation usually takes the form of a Constrained Markov Decision Process (CMDP). We focus on the case where the CMDP is unknown, and RL algorithms obtain samples to discover the model and compute an optimal constrained policy. Our goal is to characterize the relationship between safety constraints and the number of samples needed to ensure a desired level of accuracy -- both objective maximization and constraint satisfaction -- in…
University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；Sandia National Laboratories（桑迪亚国家实验室）
We consider Markov Decision Processes (MDPs) in which every stationary policy induces the same graph structure for the underlying Markov chain and further, the graph has the following property: if we replace each recurrent class by a node, then the resulting graph is acyclic. For such MDPs, we prove the convergence of the stochastic dynamics associated with a version of optimistic policy iteration (OPI), suggested in Tsitsiklis (2002), in which the values associated with all the nodes visited during each iteration of the OPI are updated.
We address the issue of safety in reinforcement learning. We pose the problem in an episodic framework of a constrained Markov decision process. Existing results have shown that it is possible to achieve a reward regret of $\tilde{\mathcal{O}}(\sqrt{K})$ while allowing an $\tilde{\mathcal{O}}(\sqrt{K})$ constraint violation in $K$ episodes. A critical question that arises is whether it is possible to keep the constraint violation even smaller. We show that when a strictly safe policy is known, then one can confine the system to zero constraint violation with arbitrarily high probability while …
Provably Efficient RL under Episode-Wise Safety in Constrained MDPs with Linear Function Approximation
The University of Tokyo（东京大学）；New Jersey Institute of Technology（新泽西理工学院）；OMRON SINIC X（欧姆龙SINIC X）；NexaScience；Kyoto University（京都大学）；Institute of Science Tokyo（东京科学大学）；DENSO IT Laboratory（电装IT实验室）
Facebook AI Research（Facebook人工智能研究院）；McGill University（麦吉尔大学）；UC Berkeley（加州大学伯克利分校）
In reinforcement learning (RL), when defining a Markov Decision Process (MDP), the environment dynamics is implicitly assumed to be stationary. This assumption of stationarity, while simplifying, can be unrealistic in many scenarios. In the continual reinforcement learning scenario, the sequence of tasks is another source of nonstationarity. In this work, we propose to examine this continual reinforcement learning setting through the block contextual MDP (BC-MDP) framework, which enables us to relax the assumption of stationarity. This framework challenges RL algorithms to handle both nonstati…
We propose a black-box reduction that turns a certain reinforcement learning algorithm with optimal regret in a (near-)stationary environment into another algorithm with optimal dynamic regret in a non-stationary environment, importantly without any prior knowledge on the degree of non-stationarity. By plugging different algorithms into our black-box, we provide a list of examples showing that our approach not only recovers recent results for (contextual) multi-armed bandits achieved by very specialized algorithms, but also significantly improves the state of the art for (generalized) linear b…
DeepMind（深度思维）；University College London（伦敦大学学院）；University of Alberta（阿尔伯塔大学）；Omron Sinic X（欧姆龙Sinic X）
We consider approximate dynamic programming in $γ$-discounted Markov decision processes and apply it to approximate planning with linear value-function approximation. Our first contribution is a new variant of Approximate Policy Iteration (API), called Confident Approximate Policy Iteration (CAPI), which computes a deterministic stationary policy with an optimal error bound scaling linearly with the product of the effective horizon $H$ and the worst-case approximation error $ε$ of the action-value functions of stationary policies. This improvement over API (whose error scales with $H^2$) comes…
Current work in explainable reinforcement learning generally produces policies in the form of a decision tree over the state space. Such policies can be used for formal safety verification, agent behavior prediction, and manual inspection of important features. However, existing approaches fit a decision tree after training or use a custom learning procedure which is not compatible with new learning techniques, such as those which use neural networks. To address this limitation, we propose a novel Markov Decision Process (MDP) type for learning decision tree policies: Iterative Bounding MDPs (…
Mitsubishi Electric Research Laboratories (MERL)（三菱电机研究实验室（MERL））
In this paper, we focus on the problem of robustifying reinforcement learning (RL) algorithms with respect to model uncertainties. Indeed, in the framework of model-based RL, we propose to merge the theory of constrained Markov decision process (CMDP), with the theory of robust Markov decision process (RMDP), leading to a formulation of robust constrained-MDPs (RCMDP). This formulation, simple in essence, allows us to design RL algorithms that are robust in performance, and provides constraint satisfaction guarantees, with respect to uncertainties in the system's states transition probabilitie…
Université de Toulouse（图卢兹大学）；ONERA - The French Aerospace Lab（法国航空航天实验室（ONERA））；ISAE-SUPAERO（法国高等航空航天学院）
This work tackles the problem of robust zero-shot planning in non-stationary stochastic environments. We study Markov Decision Processes (MDPs) evolving over time and consider Model-Based Reinforcement Learning algorithms in this setting. We make two hypotheses: 1) the environment evolves continuously with a bounded evolution rate; 2) a current model is known at each decision epoch but not its evolution. Our contribution can be presented in four points. 1) we define a specific class of MDPs that we call Non-Stationary MDPs (NSMDPs). We introduce the notion of regular evolution by making an hyp…
In many real-world applications (e.g., planetary exploration, robot navigation), an autonomous agent must be able to explore a space with guaranteed safety. Most safe exploration algorithms in the field of reinforcement learning and robotics have been based on the assumption that the safety features are a priori known and time-invariant. This paper presents a learning algorithm called ST-SafeMDP for exploring Markov decision processes (MDPs) that is based on the assumption that the safety features are a priori unknown and time-variant. In this setting, the agent explores MDPs while constrainin…
University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；Cornell University（康奈尔大学）
Monte-Carlo planning, as exemplified by Monte-Carlo Tree Search (MCTS), has demonstrated remarkable performance in applications with finite spaces. In this paper, we consider Monte-Carlo planning in an environment with continuous state-action spaces, a much less understood problem with important applications in control and robotics. We introduce POLY-HOOT, an algorithm that augments MCTS with a continuous armed bandit strategy named Hierarchical Optimistic Optimization (HOO) (Bubeck et al., 2011). Specifically, we enhance HOO by using an appropriate polynomial, rather than logarithmic, bonus t…
On the Convergence of Monte Carlo UCB for Random-Length Episodic MDPs
SFSC of AI and DL, NYU Shanghai（纽约大学上海人工智能与深度学习科学与技术中心）；New York University（纽约大学）；New York University Shanghai（纽约大学上海分校）；New York University Abu Dhabi（纽约大学阿布扎比分校）
This paper targets the efficient construction of a safety shield for decision making in scenarios that incorporate uncertainty. Markov decision processes (MDPs) are prominent models to capture such planning problems. Reinforcement learning (RL) is a machine learning technique to determine near-optimal policies in MDPs that may be unknown prior to exploring the model. However, during exploration, RL is prone to induce behavior that is undesirable or not allowed in safety- or mission-critical contexts. We introduce the concept of a probabilistic shield that enables decision-making to adhere to s…
Reinforcement Learning for Exponential Utility: Algorithms and Convergence in Discounted MDPs
Computer Science and Automation, Indian Institute of Science, Bengaluru（印度科学研究院计算机科学与自动化部门，班加罗尔）；Computer Science and Engineering, Indian Institute of Technology Madras, Chennai（印度理工学院马德拉斯分校计算机科学与工程系，钦奈）；Tata Consultancy Services Limited, Hyderabad（海得拉巴市塔塔咨询公司）
Department of Computer Science（计算机科学系）；University of New Hampshire（新罕布什尔大学）
Robust MDPs (RMDPs) can be used to compute policies with provable worst-case guarantees in reinforcement learning. The quality and robustness of an RMDP solution are determined by the ambiguity set---the set of plausible transition probabilities---which is usually constructed as a multi-dimensional confidence region. Existing methods construct ambiguity sets as confidence regions using concentration inequalities which leads to overly conservative solutions. This paper proposes a new paradigm that can achieve better solutions with the same robustness guarantees without using confidence regions …
UC Santa Barbara（加州大学圣巴巴拉分校）；Salesforce Research（Salesforce研究院）
We consider the problem of offline reinforcement learning (RL) -- a well-motivated setting of RL that aims at policy optimization using only historical data. Despite its wide applicability, theoretical understandings of offline RL, such as its optimal sample complexity, remain largely open even in basic settings such as \emph{tabular} Markov Decision Processes (MDPs). In this paper, we propose Off-Policy Double Variance Reduction (OPDVR), a new variance reduction based algorithm for offline RL. Our main result shows that OPDVR provably identifies an $ε$-optimal policy with $\widetilde{O}(H^2/d…
LTCI, Télécom Paris, Institut Polytechnique de Paris（巴黎理工学院巴黎电信学院 LTCI）；Microsoft Research Montréal（蒙特利尔微软研究院）
Batch Reinforcement Learning (Batch RL) consists in training a policy using trajectories collected with another policy, called the behavioural policy. Safe policy improvement (SPI) provides guarantees with high probability that the trained policy performs better than the behavioural policy, also called baseline in this setting. Previous work shows that the SPI objective improves mean performance as compared to using the basic RL objective, which boils down to solving the MDP with maximum likelihood. Here, we build on that work and improve more precisely the SPI with Baseline Bootstrapping algo…
Federal University of Rio Grande do Sul（里约热内卢联邦大学南里奥格兰德分校）；University of Massachusetts（马萨诸塞大学）
Non-stationary environments are challenging for reinforcement learning algorithms. If the state transition and/or reward functions change based on latent factors, the agent is effectively tasked with optimizing a behavior that maximizes performance over a possibly infinite random sequence of Markov Decision Processes (MDPs), each of which drawn from some unknown distribution. We call each such MDP a context. Most related works make strong assumptions such as knowledge about the distribution over contexts, the existence of pre-training phases, or a priori knowledge about the number, sequence, o…
We consider un-discounted reinforcement learning (RL) in Markov decision processes (MDPs) under drifting non-stationarity, i.e., both the reward and state transition distributions are allowed to evolve over time, as long as their respective total variations, quantified by suitable metrics, do not exceed certain variation budgets. We first develop the Sliding Window Upper-Confidence bound for Reinforcement Learning with Confidence Widening (SWUCRL2-CW) algorithm, and establish its dynamic regret bound when the variation budgets are known. In addition, we propose the Bandit-over-Reinforcement Le…
Robustness is important for sequential decision making in a stochastic dynamic environment with uncertain probabilistic parameters. We address the problem of using robust MDPs (RMDPs) to compute policies with provable worst-case guarantees in reinforcement learning. The quality and robustness of an RMDP solution is determined by its ambiguity set. Existing methods construct ambiguity sets that lead to impractically conservative solutions. In this paper, we propose RSVF, which achieves less conservative solutions with the same worst-case guarantees by 1) leveraging a Bayesian prior, 2) optimizi…
Graz University of Technology（格拉茨技术大学）；Silicon Austria Labs（奥地利硅实验室）；Lamarr Security Research（拉马尔安全研究院）；Aalborg University（奥尔堡大学）
Safety is still one of the major research challenges in reinforcement learning (RL). In this paper, we address the problem of how to avoid safety violations of RL agents during exploration in probabilistic and partially unknown environments. Our approach combines automata learning for Markov Decision Processes (MDPs) and shield synthesis in an iterative approach. Initially, the MDP representing the environment is unknown. The agent starts exploring the environment and collects traces. From the collected traces, we passively learn MDPs that abstractly represent the safety-relevant aspects of th…
