Although Reinforcement Learning (RL) is effective for sequential decision-making problems under uncertainty, it still fails to thrive in real-world systems where risk or safety is a binding constraint. In this paper, we formulate the RL problem with safety constraints as a non-zero-sum game. While deployed with maximum entropy RL, this formulation leads to a safe adversarially guided soft actor-critic framework, called SAAC. In SAAC, the adversary aims to break the safety constraint while the RL agent aims to maximize the constrained value function given the adversary's policy. The safety constraint on the agent's value function manifests only as a repulsion term between the agent's and the adversary's policies. Unlike previous approaches, SAAC can address different safety criteria such as safe exploration, mean-variance risk sensitivity, and CVaR-like coherent risk sensitivity. We illustrate the design of the adversary for these constraints. Then, in each of these variations, we show the agent differentiates itself from the adversary's unsafe actions in addition to learning to solve the task. Finally, for challenging continuous control tasks, we demonstrate that SAAC achieves faster convergence, better efficiency, and fewer failures to satisfy the safety constraints than risk-averse distributional RL and risk-neutral soft actor-critic algorithms.
相關內容
Deep Reinforcement Learning has demonstrated the potential of neural networks tuned with gradient descent for solving complex tasks in well-delimited environments. However, these neural systems are slow learners producing specialised agents with no mechanism to continue learning beyond their training curriculum. On the contrary, biological synaptic plasticity is persistent and manifold, and has been hypothesised to play a key role in executive functions such as working memory and cognitive flexibility, potentially supporting more efficient and generic learning abilities. Inspired by this, we propose to build networks with dynamic weights, able to continually perform self-reflexive modification as a function of their current synaptic state and action-reward feedback, rather than a fixed network configuration. The resulting model, MetODS (for Meta-Optimized Dynamical Synapses) is a broadly applicable meta-reinforcement learning system able to learn efficient and powerful control rules in the agent policy space. A single layer with dynamic synapses can perform one-shot learning, generalize navigation principles to unseen environments and demonstrate a strong ability to learn adaptive motor policies, comparing favourably with previous meta-reinforcement learning approaches.
Inverse reinforcement learning (IRL) seeks to infer a cost function that explains the underlying goals and preferences of expert demonstrations. This paper presents receding horizon inverse reinforcement learning (RHIRL), a new IRL algorithm for high-dimensional, noisy, continuous systems with black-box dynamic models. RHIRL addresses two key challenges of IRL: scalability and robustness. To handle high-dimensional continuous systems, RHIRL matches the induced optimal trajectories with expert demonstrations locally in a receding horizon manner and 'stitches' together the local solutions to learn the cost; it thereby avoids the 'curse of dimensionality'. This contrasts sharply with earlier algorithms that match with expert demonstrations globally over the entire high-dimensional state space. To be robust against imperfect expert demonstrations and system control noise, RHIRL learns a state-dependent cost function 'disentangled' from system dynamics under mild conditions. Experiments on benchmark tasks show that RHIRL outperforms several leading IRL algorithms in most instances. We also prove that the cumulative error of RHIRL grows linearly with the task duration.
Though deep reinforcement learning (DRL) has obtained substantial success, it may encounter catastrophic failures due to the intrinsic uncertainty of both transition and observation. Most of the existing methods for safe reinforcement learning can only handle transition disturbance or observation disturbance since these two kinds of disturbance affect different parts of the agent; besides, the popular worst-case return may lead to overly pessimistic policies. To address these issues, we first theoretically prove that the performance degradation under transition disturbance and observation disturbance depends on a novel metric of Value Function Range (VFR), which corresponds to the gap in the value function between the best state and the worst state. Based on the analysis, we adopt conditional value-at-risk (CVaR) as an assessment of risk and propose a novel reinforcement learning algorithm of CVaR-Proximal-Policy-Optimization (CPPO) which formalizes the risk-sensitive constrained optimization problem by keeping its CVaR under a given threshold. Experimental results show that CPPO achieves a higher cumulative reward and is more robust against both observation and transition disturbances on a series of continuous control tasks in MuJoCo.
Motivated by the recent empirical success of policy-based reinforcement learning (RL), there has been a research trend studying the performance of policy-based RL methods on standard control benchmark problems. In this paper, we examine the effectiveness of policy-based RL methods on an important robust control problem, namely $\mu$ synthesis. We build a connection between robust adversarial RL and $\mu$ synthesis, and develop a model-free version of the well-known $DK$-iteration for solving state-feedback $\mu$ synthesis with static $D$-scaling. In the proposed algorithm, the $K$ step mimics the classical central path algorithm via incorporating a recently-developed double-loop adversarial RL method as a subroutine, and the $D$ step is based on model-free finite difference approximation. Extensive numerical study is also presented to demonstrate the utility of our proposed model-free algorithm. Our study sheds new light on the connections between adversarial RL and robust control.
Recent deep reinforcement learning (DRL) successes rely on end-to-end learning from fixed-size observational inputs (e.g. image, state-variables). However, many challenging and interesting problems in decision making involve observations or intermediary representations which are best described as a set of entities: either the image-based approach would miss small but important details in the observations (e.g. ojects on a radar, vehicles on satellite images, etc.), the number of sensed objects is not fixed (e.g. robotic manipulation), or the problem simply cannot be represented in a meaningful way as an image (e.g. power grid control, or logistics). This type of structured representations is not directly compatible with current DRL architectures, however, there has been an increase in machine learning techniques directly targeting structured information, potentially addressing this issue. We propose to combine recent advances in set representations with slot attention and graph neural networks to process structured data, broadening the range of applications of DRL algorithms. This approach allows to address entity-based problems in an efficient and scalable way. We show that it can improve training time and robustness significantly, and demonstrate their potential to handle structured as well as purely visual domains, on multiple environments from the Atari Learning Environment and Simple Playgrounds.
In this work we propose Neuro-Nav, an open-source library for neurally plausible reinforcement learning (RL). RL is among the most common modeling frameworks for studying decision making, learning, and navigation in biological organisms. In utilizing RL, cognitive scientists often handcraft environments and agents to meet the needs of their particular studies. On the other hand, artificial intelligence researchers often struggle to find benchmarks for neurally and biologically plausible representation and behavior (e.g., in decision making or navigation). In order to streamline this process across both fields with transparency and reproducibility, Neuro-Nav offers a set of standardized environments and RL algorithms drawn from canonical behavioral and neural studies in rodents and humans. We demonstrate that the toolkit replicates relevant findings from a number of studies across both cognitive science and RL literatures. We furthermore describe ways in which the library can be extended with novel algorithms (including deep RL) and environments to address future research needs of the field.
Reinforcement learning has been shown to be an effective strategy for automatically training policies for challenging control problems. Focusing on non-cooperative multi-agent systems, we propose a novel reinforcement learning framework for training joint policies that form a Nash equilibrium. In our approach, rather than providing low-level reward functions, the user provides high-level specifications that encode the objective of each agent. Then, guided by the structure of the specifications, our algorithm searches over policies to identify one that provably forms an $\epsilon$-Nash equilibrium (with high probability). Importantly, it prioritizes policies in a way that maximizes social welfare across all agents. Our empirical evaluation demonstrates that our algorithm computes equilibrium policies with high social welfare, whereas state-of-the-art baselines either fail to compute Nash equilibria or compute ones with comparatively lower social welfare.
Exploration-exploitation is a powerful and practical tool in multi-agent learning (MAL), however, its effects are far from understood. To make progress in this direction, we study a smooth analogue of Q-learning. We start by showing that our learning model has strong theoretical justification as an optimal model for studying exploration-exploitation. Specifically, we prove that smooth Q-learning has bounded regret in arbitrary games for a cost model that explicitly captures the balance between game and exploration costs and that it always converges to the set of quantal-response equilibria (QRE), the standard solution concept for games under bounded rationality, in weighted potential games with heterogeneous learning agents. In our main task, we then turn to measure the effect of exploration in collective system performance. We characterize the geometry of the QRE surface in low-dimensional MAL systems and link our findings with catastrophe (bifurcation) theory. In particular, as the exploration hyperparameter evolves over-time, the system undergoes phase transitions where the number and stability of equilibria can change radically given an infinitesimal change to the exploration parameter. Based on this, we provide a formal theoretical treatment of how tuning the exploration parameter can provably lead to equilibrium selection with both positive as well as negative (and potentially unbounded) effects to system performance.
This paper surveys the field of transfer learning in the problem setting of Reinforcement Learning (RL). RL has been the key solution to sequential decision-making problems. Along with the fast advance of RL in various domains. including robotics and game-playing, transfer learning arises as an important technique to assist RL by leveraging and transferring external expertise to boost the learning process. In this survey, we review the central issues of transfer learning in the RL domain, providing a systematic categorization of its state-of-the-art techniques. We analyze their goals, methodologies, applications, and the RL frameworks under which these transfer learning techniques would be approachable. We discuss the relationship between transfer learning and other relevant topics from an RL perspective and also explore the potential challenges as well as future development directions for transfer learning in RL.
We study the problem of learning to reason in large scale knowledge graphs (KGs). More specifically, we describe a novel reinforcement learning framework for learning multi-hop relational paths: we use a policy-based agent with continuous states based on knowledge graph embeddings, which reasons in a KG vector space by sampling the most promising relation to extend its path. In contrast to prior work, our approach includes a reward function that takes the accuracy, diversity, and efficiency into consideration. Experimentally, we show that our proposed method outperforms a path-ranking based algorithm and knowledge graph embedding methods on Freebase and Never-Ending Language Learning datasets.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191