PyTorchKR Papers

최신 논문

총 1,869편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2102.03479공식 코드 · Papers with Code

Rethinking the Implementation Matters in Cooperative Multi-Agent Reinforcement Learning

Multi-Agent Reinforcement Learning (MARL) has seen revolutionary breakthroughs with its successful application to multi-agent cooperative tasks such as computer games and robot swarms. QMIX, a widely popular MARL algorithm, has been used to solve cooperative tasks, e.g. Starcraft Multi-Agent Challenge (SMAC), Difficulty-Enhanced Predator-Prey (DEPP). Recent variants of QMIX target relaxing the monotonicity constraint of QMIX, allowing for performance improvement in SMAC. However, in this paper, we investigate the code-level optimizations of these variants and the monotonicity constraint. We find that (1) such improvements of the variants are significantly affected by various code-level optimizations; (2) QMIX with normalized optimizations outperforms other previous works in SMAC; (3) the monotonicity constraint may improve sample efficiency in SMAC and DEPP. Last, a discussion with theoretical analysis is demonstrated about why QMIX works well in SMAC. We open-source the code at \url{https://github.com/hijkzzz/pymarl2}.

arXiv:2102.04402

Contrasting Centralized and Decentralized Critics in Multi-Agent Reinforcement Learning

Centralized Training for Decentralized Execution, where agents are trained offline using centralized information but execute in a decentralized manner online, has gained popularity in the multi-agent reinforcement learning community. In particular, actor-critic methods with a centralized critic and decentralized actors are a common instance of this idea. However, the implications of using a centralized critic in this context are not fully discussed and understood even though it is the standard choice of many algorithms. We therefore formally analyze centralized and decentralized critic approaches, providing a deeper understanding of the implications of critic choice. Because our theory makes unrealistic assumptions, we also empirically compare the centralized and decentralized critic methods over a wide set of environments to validate our theories and to provide practical advice. We show that there exist misconceptions regarding centralized critics in the current literature and show that the centralized critic design is not strictly beneficial, but rather both centralized and decentralized critics have different pros and cons that should be taken into account by algorithm designers.

arXiv:2102.02274

Neural Recursive Belief States in Multi-Agent Reinforcement Learning

In multi-agent reinforcement learning, the problem of learning to act is particularly difficult because the policies of co-players may be heavily conditioned on information only observed by them. On the other hand, humans readily form beliefs about the knowledge possessed by their peers and leverage beliefs to inform decision-making. Such abilities underlie individual success in a wide range of Markov games, from bluffing in Poker to conditional cooperation in the Prisoner's Dilemma, to convention-building in Bridge. Classical methods are usually not applicable to complex domains due to the intractable nature of hierarchical beliefs (i.e. beliefs of other agents' beliefs). We propose a scalable method to approximate these belief structures using recursive deep generative models, and to use the belief models to obtain representations useful to acting in complex tasks. Our agents trained with belief models outperform model-free baselines with equivalent representational capacity using common training paradigms. We also show that higher-order belief models outperform agents with lower-order models.

arXiv:2102.01434

An Abstraction-based Method to Check Multi-Agent Deep Reinforcement-Learning Behaviors

Multi-agent reinforcement learning (RL) often struggles to ensure the safe behaviours of the learning agents, and therefore it is generally not adapted to safety-critical applications. To address this issue, we present a methodology that combines formal verification with (deep) RL algorithms to guarantee the satisfaction of formally-specified safety constraints both in training and testing. The approach we propose expresses the constraints to verify in Probabilistic Computation Tree Logic (PCTL) and builds an abstract representation of the system to reduce the complexity of the verification step. This abstract model allows for model checking techniques to identify a set of abstract policies that meet the safety constraints expressed in PCTL. Then, the agents' behaviours are restricted according to these safe abstract policies. We provide formal guarantees that by using this method, the actions of the agents always meet the safety constraints, and provide a procedure to generate an abstract model automatically. We empirically evaluate and show the effectiveness of our method in a multi-agent environment.

arXiv:2102.01611

Towards Multi-agent Reinforcement Learning for Wireless Network Protocol Synthesis

This paper proposes a multi-agent reinforcement learning based medium access framework for wireless networks. The access problem is formulated as a Markov Decision Process (MDP), and solved using reinforcement learning with every network node acting as a distributed learning agent. The solution components are developed step by step, starting from a single-node access scenario in which a node agent incrementally learns to control MAC layer packet loads for reining in self-collisions. The strategy is then scaled up for multi-node fully-connected scenarios by using more elaborate reward structures. It also demonstrates preliminary feasibility for more general partially connected topologies. It is shown that by learning to adjust MAC layer transmission probabilities, the protocol is not only able to attain theoretical maximum throughput at an optimal load, but unlike classical approaches, it can also retain that maximum throughput at higher loading conditions. Additionally, the mechanism is agnostic to heterogeneous loading while preserving that feature. It is also shown that access priorities of the protocol across nodes can be parametrically adjusted. Finally, it is also shown that the online learning feature of reinforcement learning is able to make the protocol adapt to time-varying loading conditions.

arXiv:2102.00824공식 코드 · Papers with Code

HAMMER: Multi-Level Coordination of Reinforcement Learning Agents via Learned Messaging

Cooperative multi-agent reinforcement learning (MARL) has achieved significant results, most notably by leveraging the representation-learning abilities of deep neural networks. However, large centralized approaches quickly become infeasible as the number of agents scale, and fully decentralized approaches can miss important opportunities for information sharing and coordination. Furthermore, not all agents are equal -- in some cases, individual agents may not even have the ability to send communication to other agents or explicitly model other agents. This paper considers the case where there is a single, powerful, \emph{central agent} that can observe the entire observation space, and there are multiple, low-powered \emph{local agents} that can only receive local observations and are not able to communicate with each other. The central agent's job is to learn what message needs to be sent to different local agents based on the global observations, not by centrally solving the entire problem and sending action commands, but by determining what additional information an individual agent should receive so that it can make a better decision. In this work we present our MARL algorithm \algo, describe where it would be most applicable, and implement it in the cooperative navigation and multi-agent walker domains. Empirical results show that 1) learned communication does indeed improve system performance, 2) results generalize to heterogeneous local agents, and 3) results generalize to different reward structures.

arXiv:2102.00582공식 코드 · Papers with Code

Multi-Agent Reinforcement Learning with Temporal Logic Specifications

In this paper, we study the problem of learning to satisfy temporal logic specifications with a group of agents in an unknown environment, which may exhibit probabilistic behaviour. From a learning perspective these specifications provide a rich formal language with which to capture tasks or objectives, while from a logic and automated verification perspective the introduction of learning capabilities allows for practical applications in large, stochastic, unknown environments. The existing work in this area is, however, limited. Of the frameworks that consider full linear temporal logic or have correctness guarantees, all methods thus far consider only the case of a single temporal logic specification and a single agent. In order to overcome this limitation, we develop the first multi-agent reinforcement learning technique for temporal logic specifications, which is also novel in its ability to handle multiple specifications. We provide correctness and convergence guarantees for our main algorithm - ALMANAC (Automaton/Logic Multi-Agent Natural Actor-Critic) - even when using function approximation. Alongside our theoretical results, we further demonstrate the applicability of our technique via a set of preliminary experiments.

arXiv:2102.01004

Hybrid Information-driven Multi-agent Reinforcement Learning

Information theoretic sensor management approaches are an ideal solution to state estimation problems when considering the optimal control of multi-agent systems, however they are too computationally intensive for large state spaces, especially when considering the limited computational resources typical of large-scale distributed multi-agent systems. Reinforcement learning (RL) is a promising alternative which can find approximate solutions to distributed optimal control problems that take into account the resource constraints inherent in many systems of distributed agents. However, the RL training can be prohibitively inefficient, especially in low-information environments where agents receive little to no feedback in large portions of the state space. We propose a hybrid information-driven multi-agent reinforcement learning (MARL) approach that utilizes information theoretic models as heuristics to help the agents navigate large sparse state spaces, coupled with information based rewards in an RL framework to learn higher-level policies. This paper presents our ongoing work towards this objective. Our preliminary findings show that such an approach can result in a system of agents that are approximately three orders of magnitude more efficient at exploring a sparse state space than naive baseline metrics. While the work is still in its early stages, it provides a promising direction for future research.

arXiv:2101.11196

Safe Multi-Agent Reinforcement Learning via Shielding

Multi-agent reinforcement learning (MARL) has been increasingly used in a wide range of safety-critical applications, which require guaranteed safety (e.g., no unsafe states are ever visited) during the learning process.Unfortunately, current MARL methods do not have safety guarantees. Therefore, we present two shielding approaches for safe MARL. In centralized shielding, we synthesize a single shield to monitor all agents' joint actions and correct any unsafe action if necessary. In factored shielding, we synthesize multiple shields based on a factorization of the joint state space observed by all agents; the set of shields monitors agents concurrently and each shield is only responsible for a subset of agents at each step.Experimental results show that both approaches can guarantee the safety of agents during learning without compromising the quality of learned policies; moreover, factored shielding is more scalable in the number of agents than centralized shielding.

arXiv:2101.10721공식 코드 · Papers with Code

Data sharing games

Data sharing issues pervade online social and economic environments. To foster social progress, it is important to develop models of the interaction between data producers and consumers that can promote the rise of cooperation between the involved parties. We formalize this interaction as a game, the data sharing game, based on the Iterated Prisoner's Dilemma and deal with it through multi-agent reinforcement learning techniques. We consider several strategies for how the citizens may behave, depending on the degree of centralization sought. Simulations suggest mechanisms for cooperation to take place and, thus, achieve maximum social utility: data consumers should perform some kind of opponent modeling, or a regulator should transfer utility between both players and incentivise them.

arXiv:2101.10369

Effective Communications: A Joint Learning and Communication Framework for Multi-Agent Reinforcement Learning over Noisy Channels

We propose a novel formulation of the "effectiveness problem" in communications, put forth by Shannon and Weaver in their seminal work [2], by considering multiple agents communicating over a noisy channel in order to achieve better coordination and cooperation in a multi-agent reinforcement learning (MARL) framework. Specifically, we consider a multi-agent partially observable Markov decision process (MA-POMDP), in which the agents, in addition to interacting with the environment can also communicate with each other over a noisy communication channel. The noisy communication channel is considered explicitly as part of the dynamics of the environment and the message each agent sends is part of the action that the agent can take. As a result, the agents learn not only to collaborate with each other but also to communicate "effectively" over a noisy channel. This framework generalizes both the traditional communication problem, where the main goal is to convey a message reliably over a noisy channel, and the "learning to communicate" framework that has received recent attention in the MARL literature, where the underlying communication channels are assumed to be error-free. We show via examples that the joint policy learned using the proposed framework is superior to that where the communication is considered separately from the underlying MA-POMDP. This is a very powerful framework, which has many real world applications, from autonomous vehicle planning to drone swarm control, and opens up the rich toolbox of deep reinforcement learning for the design of multi-user communication systems.

arXiv:2101.09698

Fast Sequence Generation with Multi-Agent Reinforcement Learning

Autoregressive sequence Generation models have achieved state-of-the-art performance in areas like machine translation and image captioning. These models are autoregressive in that they generate each word by conditioning on previously generated words, which leads to heavy latency during inference. Recently, non-autoregressive decoding has been proposed in machine translation to speed up the inference time by generating all words in parallel. Typically, these models use the word-level cross-entropy loss to optimize each word independently. However, such a learning process fails to consider the sentence-level consistency, thus resulting in inferior generation quality of these non-autoregressive models. In this paper, we propose a simple and efficient model for Non-Autoregressive sequence Generation (NAG) with a novel training paradigm: Counterfactuals-critical Multi-Agent Learning (CMAL). CMAL formulates NAG as a multi-agent reinforcement learning system where element positions in the target sequence are viewed as agents that learn to cooperatively maximize a sentence-level reward. On MSCOCO image captioning benchmark, our NAG method achieves a performance comparable to state-of-the-art autoregressive models, while brings 13.9x decoding speedup. On WMT14 EN-DE machine translation dataset, our method outperforms cross-entropy trained baseline by 6.0 BLEU points while achieves the greatest decoding speedup of 17.46x.

arXiv:2101.08001공식 코드 · Papers with Code

UPDeT: Universal Multi-agent Reinforcement Learning via Policy Decoupling with Transformers

Recent advances in multi-agent reinforcement learning have been largely limited in training one model from scratch for every new task. The limitation is due to the restricted model architecture related to fixed input and output dimensions. This hinders the experience accumulation and transfer of the learned agent over tasks with diverse levels of difficulty (e.g. 3 vs 3 or 5 vs 6 multi-agent games). In this paper, we make the first attempt to explore a universal multi-agent reinforcement learning pipeline, designing one single architecture to fit tasks with the requirement of different observation and action configurations. Unlike previous RNN-based models, we utilize a transformer-based model to generate a flexible policy by decoupling the policy distribution from the intertwined input observation with an importance weight measured by the merits of the self-attention mechanism. Compared to a standard transformer block, the proposed model, named as Universal Policy Decoupling Transformer (UPDeT), further relaxes the action restriction and makes the multi-agent task's decision process more explainable. UPDeT is general enough to be plugged into any multi-agent reinforcement learning pipeline and equip them with strong generalization abilities that enables the handling of multiple tasks at a time. Extensive experiments on large-scale SMAC multi-agent competitive games demonstrate that the proposed UPDeT-based multi-agent reinforcement learning achieves significant results relative to state-of-the-art approaches, demonstrating advantageous transfer capability in terms of both performance and training speed (10 times faster).

arXiv:2101.06890

Cooperative and Competitive Biases for Multi-Agent Reinforcement Learning

Training a multi-agent reinforcement learning (MARL) algorithm is more challenging than training a single-agent reinforcement learning algorithm, because the result of a multi-agent task strongly depends on the complex interactions among agents and their interactions with a stochastic and dynamic environment. We propose an algorithm that boosts MARL training using the biased action information of other agents based on a friend-or-foe concept. For a cooperative and competitive environment, there are generally two groups of agents: cooperative-agents and competitive-agents. In the proposed algorithm, each agent updates its value function using its own action and the biased action information of other agents in the two groups. The biased joint action of cooperative agents is computed as the sum of their actual joint action and the imaginary cooperative joint action, by assuming all the cooperative agents jointly maximize the target agent's value function. The biased joint action of competitive agents can be computed similarly. Each agent then updates its own value function using the biased action information, resulting in a biased value function and corresponding biased policy. Subsequently, the biased policy of each agent is inevitably subjected to recommend an action to cooperate and compete with other agents, thereby introducing more active interactions among agents and enhancing the MARL policy learning. We empirically demonstrate that our algorithm outperforms existing algorithms in various mixed cooperative-competitive environments. Furthermore, the introduced biases gradually decrease as the training proceeds and the correction based on the imaginary assumption vanishes.

논문 ID: 8761.33988

Multirobot Coverage of Modular Environments

Multirobot systems for covering environments are increasingly used in applications like cleaning, industrial inspection, patrolling, and precision agriculture. The problem of covering a given environment using multiple robots can be naturally formulated and studied as a multi-Traveling Salesperson Problem (mTSP). In a mTSP, the environment is represented as a graph and the goal is to find tours (starting and ending at the same depot) for the robots in order to visit all the vertices with minimum global cost, which is typically calculated as the makespan, namely the length of the longest tour. The mTSP is an NP-hard problem for which several approximation algorithms have been proposed. These algorithms usually assume generic environments, but tighter approximation bounds can be reached focusing on specific environments. In this paper, we address the case of modular environments, namely of environments composed of sub-parts, called modules, that can be reached from each other only through some linking structures. Examples are multi-floor buildings, in which the modules are the floors and the linking structures are the staircases or the elevators, and floors of large hotels or hospitals, in which the modules are the rooms and the linking structures are the corridors. We focus on linear modular environments, with the modules organized sequentially, presenting an efficient (with polynomial worst-case time complexity) algorithm that finds a solution for the mTSP whose cost is within a bounded distance from the cost of the optimal solution. The main idea of our algorithm is to allocate disjoint "blocks'' of adjacent modules to the robots, in such a way that each module is covered by only one robot. We experimentally compare our algorithm against some state-of-the-art algorithms for solving mTSPs in generic environments and show that it is able to provide solutions with lower makespan and spending a computing time several orders of magnitude shorter.

arXiv:2101.04237공식 코드 · Papers with Code

Solving Common-Payoff Games with Approximate Policy Iteration

For artificially intelligent learning systems to have widespread applicability in real-world settings, it is important that they be able to operate decentrally. Unfortunately, decentralized control is difficult -- computing even an epsilon-optimal joint policy is a NEXP complete problem. Nevertheless, a recently rediscovered insight -- that a team of agents can coordinate via common knowledge -- has given rise to algorithms capable of finding optimal joint policies in small common-payoff games. The Bayesian action decoder (BAD) leverages this insight and deep reinforcement learning to scale to games as large as two-player Hanabi. However, the approximations it uses to do so prevent it from discovering optimal joint policies even in games small enough to brute force optimal solutions. This work proposes CAPI, a novel algorithm which, like BAD, combines common knowledge with deep reinforcement learning. However, unlike BAD, CAPI prioritizes the propensity to discover optimal joint policies over scalability. While this choice precludes CAPI from scaling to games as large as Hanabi, empirical results demonstrate that, on the games to which CAPI does scale, it is capable of discovering optimal joint policies even when other modern multi-agent reinforcement learning algorithms are unable to do so. Code is available at https://github.com/ssokota/capi .

arXiv:2101.02308

Coding for Distributed Multi-Agent Reinforcement Learning

This paper aims to mitigate straggler effects in synchronous distributed learning for multi-agent reinforcement learning (MARL) problems. Stragglers arise frequently in a distributed learning system, due to the existence of various system disturbances such as slow-downs or failures of compute nodes and communication bottlenecks. To resolve this issue, we propose a coded distributed learning framework, which speeds up the training of MARL algorithms in the presence of stragglers, while maintaining the same accuracy as the centralized approach. As an illustration, a coded distributed version of the multi-agent deep deterministic policy gradient(MADDPG) algorithm is developed and evaluated. Different coding schemes, including maximum distance separable (MDS)code, random sparse code, replication-based code, and regular low density parity check (LDPC) code are also investigated. Simulations in several multi-robot problems demonstrate the promising performance of the proposed framework.

arXiv:2101.01041

Derivative-Free Policy Optimization for Linear Risk-Sensitive and Robust Control Design: Implicit Regularization and Sample Complexity

Direct policy search serves as one of the workhorses in modern reinforcement learning (RL), and its applications in continuous control tasks have recently attracted increasing attention. In this work, we investigate the convergence theory of policy gradient (PG) methods for learning the linear risk-sensitive and robust controller. In particular, we develop PG methods that can be implemented in a derivative-free fashion by sampling system trajectories, and establish both global convergence and sample complexity results in the solutions of two fundamental settings in risk-sensitive and robust control: the finite-horizon linear exponential quadratic Gaussian, and the finite-horizon linear-quadratic disturbance attenuation problems. As a by-product, our results also provide the first sample complexity for the global convergence of PG methods on solving zero-sum linear-quadratic dynamic games, a nonconvex-nonconcave minimax optimization problem that serves as a baseline setting in multi-agent reinforcement learning (MARL) with continuous spaces. One feature of our algorithms is that during the learning phase, a certain level of robustness/risk-sensitivity of the controller is preserved, which we termed as the implicit regularization property, and is an essential requirement in safety-critical control systems.

arXiv:2101.00746공식 코드 · Papers with Code

MetaVIM: Meta Variationally Intrinsic Motivated Reinforcement Learning for Decentralized Traffic Signal Control

Traffic signal control aims to coordinate traffic signals across intersections to improve the traffic efficiency of a district or a city. Deep reinforcement learning (RL) has been applied to traffic signal control recently and demonstrated promising performance where each traffic signal is regarded as an agent. However, there are still several challenges that may limit its large-scale application in the real world. To make the policy learned from a training scenario generalizable to new unseen scenarios, a novel Meta Variationally Intrinsic Motivated (MetaVIM) RL method is proposed to learn the decentralized policy for each intersection that considers neighbor information in a latent way. Specifically, we formulate the policy learning as a meta-learning problem over a set of related tasks, where each task corresponds to traffic signal control at an intersection whose neighbors are regarded as the unobserved part of the state. Then, a learned latent variable is introduced to represent the task's specific information and is further brought into the policy for learning. In addition, to make the policy learning stable, a novel intrinsic reward is designed to encourage each agent's received rewards and observation transition to be predictable only conditioned on its own history. Extensive experiments conducted on CityFlow demonstrate that the proposed method substantially outperforms existing approaches and shows superior generalizability.

arXiv:2012.15472공식 코드 · Papers with Code

Multi-Agent Reinforcement Learning for Unmanned Aerial Vehicle Coordination by Multi-Critic Policy Gradient Optimization

Recent technological progress in the development of Unmanned Aerial Vehicles (UAVs) together with decreasing acquisition costs make the application of drone fleets attractive for a wide variety of tasks. In agriculture, disaster management, search and rescue operations, commercial and military applications, the advantage of applying a fleet of drones originates from their ability to cooperate autonomously. Multi-Agent Reinforcement Learning approaches that aim to optimize a neural network based control policy, such as the best performing actor-critic policy gradient algorithms, struggle to effectively back-propagate errors of distinct rewards signal sources and tend to favor lucrative signals while neglecting coordination and exploitation of previously learned similarities. We propose a Multi-Critic Policy Optimization architecture with multiple value estimating networks and a novel advantage function that optimizes a stochastic actor policy network to achieve optimal coordination of agents. Consequently, we apply the algorithm to several tasks that require the collaboration of multiple drones in a physics-based reinforcement learning environment. Our approach achieves a stable policy network update and similarity in reward signal development for an increasing number of agents. The resulting policy achieves optimal coordination and compliance with constraints such as collision avoidance.