Generating novel graph structures that optimize given objectives while obeying some given underlying rules is fundamental for chemistry, biology and social science research. This is especially important in the task of molecular graph generation, whose goal is to discover novel molecules with desired properties such as drug-likeness and synthetic accessibility, while obeying physical laws such as chemical valency. However, designing models to find molecules that optimize desired properties while incorporating highly complex and non-differentiable rules remains to be a challenging task. Here we propose Graph Convolutional Policy Network (GCPN), a general graph convolutional network based model for goal-directed graph generation through reinforcement learning. The model is trained to optimize domain-specific rewards and adversarial loss through policy gradient, and acts in an environment that incorporates domain-specific rules. Experimental results show that GCPN can achieve 61% improvement on chemical property optimization over state-of-the-art baselines while resembling known molecules, and achieve 184% improvement on the constrained property optimization task.
Many currently deployed Reinforcement Learning agents work in an environment shared with humans, be them co-workers, users or clients. It is desirable that these agents adjust to people's preferences, learn faster thanks to their help, and act safely around them. We argue that most current approaches that learn from human feedback are unsafe: rewarding or punishing the agent a-posteriori cannot immediately prevent it from wrong-doing. In this paper, we extend Policy Gradient to make it robust to external directives, that would otherwise break the fundamentally on-policy nature of Policy Gradie…
Microsoft Research（微软研究院）；Facebook AI Research（Facebook 人工智能研究院）；University of Washington（华盛顿大学）；Cornell University（康奈尔大学）
Direct policy gradient methods for reinforcement learning are a successful approach for a variety of reasons: they are model free, they directly optimize the performance metric of interest, and they allow for richly parameterized policies. Their primary drawback is that, by being local in nature, they fail to adequately explore the environment. In contrast, while model-based approaches and Q-learning directly handle exploration through the use of optimism, their ability to handle model misspecification and function approximation is far less evident. This work introduces the the Policy Cover-Po…
Meta Reinforcement Learning (RL) methods focus on automating the design of RL algorithms that generalize to a wide range of environments. The framework introduced in (Anonymous, 2020) addresses the problem by representing different RL algorithms as Directed Acyclic Graphs (DAGs), and using an evolutionary meta learner to modify these graphs and find good agent update rules. While the search language used to generate graphs in the paper serves to represent numerous already-existing RL algorithms (e.g., DQN, DDQN), it has limitations when it comes to representing Policy Gradient algorithms. In t…
Efficient exploration is necessary to achieve good sample efficiency for reinforcement learning in general. From small, tabular settings such as gridworlds to large, continuous and sparse reward settings such as robotic object manipulation tasks, exploration through adding an uncertainty bonus to the reward function has been shown to be effective when the uncertainty is able to accurately drive exploration towards promising states. However reward bonuses can still be inefficient since they are non-stationary, which means that we must wait for function approximators to catch up and converge aga…
Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所）；DeepMind
Standard planners for sequential decision making (including Monte Carlo planning, tree search, dynamic programming, etc.) are constrained by an implicit sequential planning assumption: The order in which a plan is constructed is the same in which it is executed. We consider alternatives to this assumption for the class of goal-directed Reinforcement Learning (RL) problems. Instead of an environment transition model, we assume an imperfect, goal-directed policy. This low-level policy can be improved by a plan, consisting of an appropriate sequence of sub-goals that guide it from the start to th…
Indian Institute of Technology Madras（印度理工学院马德拉斯分校）
Sparse reward problems are one of the biggest challenges in Reinforcement Learning. Goal-directed tasks are one such sparse reward problems where a reward signal is received only when the goal is reached. One promising way to train an agent to perform goal-directed tasks is to use Hindsight Learning approaches. In these approaches, even when an agent fails to reach the desired goal, the agent learns to reach the goal it achieved instead. Doing this over multiple trajectories while generalizing the policy learned from the achieved goals, the agent learns a goal conditioned policy to reach any g…
We study an exploration method for model-free RL that generalizes the counter-based exploration bonus methods and takes into account long term exploratory value of actions rather than a single step look-ahead. We propose a model-free RL method that modifies Delayed Q-learning and utilizes the long-term exploration bonus with provable efficiency. We show that our proposed method finds a near-optimal policy in polynomial time (PAC-MDP), and also provide experimental evidence that our proposed algorithm is an efficient exploration method.
Skolkovo Institute of Science and Technology（斯科尔科沃科学技术研究院）
We propose a Reinforcement Learning based approach to approximately solve the Tree Decomposition (TD) problem. TD is a combinatorial problem, which is central to the analysis of graph minor structure and computational complexity, as well as in the algorithms of probabilistic inference, register allocation, and other practical tasks. Recently, it has been shown that combinatorial problems can be successively solved by learned heuristics. However, the majority of existing works do not address the question of the generalization of learning-based solutions. Our model is based on the graph convolut…
North Carolina State University（北卡罗来纳州立大学）；Oklahoma State University（俄克拉荷马州立大学）；DEVCOM U.S. Army Research Laboratory（美国陆军作战能力发展司令部陆军研究实验室）
Existing distributed cooperative multi-agent reinforcement learning (MARL) frameworks usually assume undirected coordination graphs and communication graphs while estimating a global reward via consensus algorithms for policy evaluation. Such a framework may induce expensive communication costs and exhibit poor scalability due to requirement of global consensus. In this work, we study MARLs with directed coordination graphs, and propose a distributed RL algorithm where the local policy evaluations are based on local value functions. The local value function of each agent is obtained by local c…
We consider the off-policy evaluation problem of reinforcement learning using deep convolutional neural networks. We analyze the deep fitted Q-evaluation method for estimating the expected cumulative reward of a target policy, when the data are generated from an unknown behavior policy. We show that, by choosing network size appropriately, one can leverage any low-dimensional manifold structure in the Markov decision process and obtain a sample-efficient estimator without suffering from the curse of high data ambient dimensionality. Specifically, we establish a sharp error bound for fitted Q-e…
Despite the close connection between exploration and sample efficiency, most state of the art reinforcement learning algorithms include no considerations for exploration beyond maximizing the entropy of the policy. In this work we address this seeming missed opportunity. We observe that the most common formulation of directed exploration in deep RL, known as bonus-based exploration (BBE), suffers from bias and slow coverage in the few-sample regime. This causes BBE to be actively detrimental to policy learning in many control tasks. We show that by decoupling the task policy from the explorati…
Reinforcement Learning with a Bilevel World-Model Architecture for Scan-Order Optimisation in Laser Directed Energy Deposition
College of Engineering, Design and Physical Sciences, Brunel University London（布鲁内尔大学伦敦工程、设计与物理科学学院）；Pattern Recognition Laboratory, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所模式识别实验室）；ISIS Neutron and Muon Source, Science and Technology Facilities Council, Rutherford Appleton Laboratory（Rutherford Appleton实验室，科学与技术设施委员会ISIS中子与μ子源）
Key Laboratory of Intelligent Information Processing of Chinese Academy of Sciences (CAS), Institute of Computing Technology, CAS（中国科学院计算技术研究所中国科学院智能信息处理重点实验室）；University of Chinese Academy of Sciences（中国科学院大学）
Lip-reading aims to infer the speech content from the lip movement sequence and can be seen as a typical sequence-to-sequence (seq2seq) problem which translates the input image sequence of lip movements to the text sequence of the speech content. However, the traditional learning process of seq2seq models always suffers from two problems: the exposure bias resulted from the strategy of "teacher-forcing", and the inconsistency between the discriminative optimization target (usually the cross-entropy loss) and the final evaluation metric (usually the character/word error rate). In this paper, we…
Being able to reach any desired location in the environment can be a valuable asset for an agent. Learning a policy to navigate between all pairs of states individually is often not feasible. An all-goals updating algorithm uses each transition to learn Q-values towards all goals simultaneously and off-policy. However the expensive numerous updates in parallel limited the approach to small tabular cases so far. To tackle this problem we propose to use convolutional network architectures to generate Q-values and updates for a large number of goals at once. We demonstrate the accuracy and genera…
Reinforcement Learning (RL) is a general framework concerned with an agent that seeks to maximize rewards in an environment. The learning typically happens through trial and error using explorative methods, such as epsilon-greedy. There are two approaches, model-based and model-free reinforcement learning, that show concrete results in several disciplines. Model-based RL learns a model of the environment for learning the policy while model-free approaches are fully explorative and exploitative without considering the underlying environment dynamics. Model-free RL works conceptually well in sim…
University of Pennsylvania（宾夕法尼亚大学）；Georgia Institute of Technology（佐治亚理工学院）
The advent of large pre-trained generative language models has provided a common framework for AI story generation via sampling the model to create sequences that continue the story. However, sampling alone is insufficient for story generation. In particular, it is hard to direct a language model to create stories to reach a specific goal event. We present two automated techniques grounded in deep reinforcement learning and reward shaping to control the plot of computer-generated stories. The first utilizes proximal policy optimization to fine-tune an existing transformer-based language model …
C-IDS: Solving Contextual POMDP via Information-Directed Objective
Department of Electrical and Computer Engineering, University of Florida, Gainesville, Florida 32608, USA（电气与计算机工程系，佛罗里达大学，盖恩斯维尔，佛罗里达州32608，美国）；Aerospace Engineering, DEVCOM Army Research Laboratory, Adelphi, MD 21001, USA（航空航天工程，DEVCOM陆军研究实验室，阿德里菲，马里兰州21001，美国）
We study the problem of explaining a rich class of behavioral properties of deep neural networks. Distinctively, our influence-directed explanations approach this problem by peering inside the network to identify neurons with high influence on a quantity and distribution of interest, using an axiomatically-justified influence measure, and then providing an interpretation for the concepts these neurons represent. We evaluate our approach by demonstrating a number of its unique capabilities on convolutional neural networks trained on ImageNet. Our evaluation demonstrates that influence-directed …
INRIA Bordeaux Sud-Ouest（法国国家信息与自动化研究所波尔多西南中心）；Sorbonne Université（索邦大学）；CNRS（法国国家科学研究中心）；German Aerospace Center (DLR)（德国航空航天中心）
Continuous action policy search is currently the focus of intensive research, driven both by the recent success of deep reinforcement learning algorithms and the emergence of competitors based on evolutionary algorithms. In this paper, we present a broad survey of policy search methods, providing a unified perspective on very different approaches, including also Bayesian Optimization and directed exploration methods. The main message of this overview is in the relationship between the families of methods, but we also outline some factors underlying sample efficiency properties of the various a…
This paper presents a novel form of policy gradient for model-free reinforcement learning (RL) with improved exploration properties. Current policy-based methods use entropy regularization to encourage undirected exploration of the reward landscape, which is ineffective in high dimensional spaces with sparse rewards. We propose a more directed exploration strategy that promotes exploration of under-appreciated reward regions. An action sequence is considered under-appreciated if its log-probability under the current policy under-estimates its resulting reward. The proposed exploration strategy…
Directed exploration strategies for reinforcement learning are critical for learning an optimal policy in a minimal number of interactions with the environment. Many algorithms use optimism to direct exploration, either through visitation estimates or upper confidence bounds, as opposed to data-inefficient strategies like ε-greedy that use random, undirected exploration. Most data-efficient exploration methods require significant computation, typically relying on a learned model to guide exploration. Least-squares methods have the potential to provide some of the data-efficiency benefits of mo…
University of Bologna（博洛尼亚大学）；Politecnico di Milano（米兰理工大学）
Uncertainty quantification has been extensively used as a means to achieve efficient directed exploration in Reinforcement Learning (RL). However, state-of-the-art methods for continuous actions still suffer from high sample complexity requirements. Indeed, they either completely lack strategies for propagating the epistemic uncertainty throughout the updates, or they mix it with aleatoric uncertainty while learning the full return distribution (e.g., distributional RL). In this paper, we propose Wasserstein Actor-Critic (WAC), an actor-critic architecture inspired by the recent Wasserstein Q-…
Information-directed sampling for bandits: a primer
Institute of Physics, School of Basic Sciences, École Polytechnique Fédérale de Lausanne - EPFL, Lausanne, Switzerland；Quantitative Life Sciences section, The Abdus Salam International Center for Theoretical Physics (ICTP), Trieste, Italy；Department of Oncology, Universit\`a degli Studi di Torino, Italy
University of British Columbia（不列颠哥伦比亚大学）；University of Toronto（多伦多大学）
We consider the problem belief-state monitoring for the purposes of implementing a policy for a partially-observable Markov decision process (POMDP), specifically how one might approximate the belief state. Other schemes for belief-state approximation (e.g., based on minimixing a measures such as KL-diveregence between the true and estimated state) are not necessarily appropriate for POMDPs. Instead we propose a framework for analyzing value-directed approximation schemes, where approximation quality is determined by the expected error in utility rather than by the error in the belief state it…
We consider the problem of approximate belief-state monitoring using particle filtering for the purposes of implementing a policy for a partially-observable Markov decision process (POMDP). While particle filtering has become a widely-used tool in AI for monitoring dynamical systems, rather scant attention has been paid to their use in the context of decision making. Assuming the existence of a value function, we derive error bounds on decision quality associated with filtering using importance sampling. We also describe an adaptive procedure that can be used to dynamically determine the numbe…
Graph convolutional networks(GCNs) have become the most popular approaches for graph data in these days because of their powerful ability to extract features from graph. GCNs approaches are divided into two categories, spectral-based and spatial-based. As the earliest convolutional networks for graph data, spectral-based GCNs have achieved impressive results in many graph related analytics tasks. However, spectral-based models cannot directly work on directed graphs. In this paper, we propose an improved spectral-based GCN for the directed graph by leveraging redefined Laplacians to improve it…
Graph Convolutional Networks (GCNs) have been widely used due to their outstanding performance in processing graph-structured data. However, the undirected graphs limit their application scope. In this paper, we extend spectral-based graph convolution to directed graphs by using first- and second-order proximity, which can not only retain the connection properties of the directed graph, but also expand the receptive field of the convolution operation. A new GCN model, called DGCN, is then designed to learn representations on the directed graph, leveraging both the first- and second-order proxi…
King Abdullah University of Science and Technology（阿卜杜拉国王科技大学）
We introduce optimization methods for convolutional neural networks that can be used to improve existing gradient-based optimization in terms of generalization error. The method requires only simple processing of existing stochastic gradients, can be used in conjunction with any optimizer, and has only a linear overhead (in the number of parameters) compared to computation of the stochastic gradient. The method works by computing the gradient of the loss function with respect to output-channel directed re-weighted L2 or Sobolev metrics, which has the effect of smoothing components of the gradi…
GRAIL: Goal Recognition Alignment through Imitation Learning
Department of Computer Science, Bar Ilan University（巴伊兰大学计算机科学系）；Department of Computer Science, Aberdeen University（阿伯丁大学计算机科学系）；Department of Computer Science, Tufts University（塔夫茨大学计算机科学系）
Clemson University（克莱姆森大学）；Université de Montréal（蒙特利尔大学）；Accenture (United States)（埃森哲（美国））；Arizona State University（亚利桑那州立大学）；Northeastern University（东北大学）
ETH Zurich（苏黎世联邦理工学院）；University of Vienna（维也纳大学）；Microsoft Research Cambridge（微软剑桥研究院）；Spotify（声田）
For many reinforcement learning (RL) applications, specifying a reward is difficult. This paper considers an RL setting where the agent obtains information about the reward only by querying an expert that can, for example, evaluate individual states or provide binary preferences over trajectories. From such expensive feedback, we aim to learn a model of the reward that allows standard RL algorithms to achieve high expected returns with as few expert queries as possible. To this end, we propose Information Directed Reward Learning (IDRL), which uses a Bayesian model of the reward and selects qu…
Tsinghua University（清华大学）；University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
This paper proposes a \emph{fully asynchronous} scheme for the policy evaluation problem of distributed reinforcement learning (DisRL) over directed peer-to-peer networks. Without waiting for any other node of the network, each node can locally update its value function at any time by using (possibly delayed) information from its neighbors. This is in sharp contrast to the gossip-based scheme where a pair of nodes concurrently update. Though the fully asynchronous setting involves a difficult multi-timescale decision problem, we design a novel stochastic average gradient (SAG) based distribute…
University of California, Santa Barbara（加州大学圣巴巴拉分校）
Modern vision-based reinforcement learning techniques often use convolutional neural networks (CNN) as universal function approximators to choose which action to take for a given visual input. Until recently, CNNs have been treated like black-box functions, but this mindset is especially dangerous when used for control in safety-critical settings. In this paper, we present our extensions of CNN visualization algorithms to the domain of vision-based reinforcement learning. We use a simulated drone environment as an example scenario. These visualization algorithms are an important tool for behav…
Efficient exploration is a crucial challenge in deep reinforcement learning. Several methods, such as behavioral priors, are able to leverage offline data in order to efficiently accelerate reinforcement learning on complex tasks. However, if the task at hand deviates excessively from the demonstrated task, the effectiveness of such methods is limited. In our work, we propose to learn features from offline data that are shared by a more diverse range of tasks, such as correlation between actions and directedness. Therefore, we introduce state-free priors, which directly model temporal consiste…
King’s College London（伦敦国王学院）；University of Oxford（牛津大学）
Off-policy reinforcement learning (RL) from pixel observations is notoriously unstable. As a result, many successful algorithms must combine different domain-specific practices and auxiliary losses to learn meaningful behaviors in complex environments. In this work, we provide novel analysis demonstrating that these instabilities arise from performing temporal-difference learning with a convolutional encoder and low-magnitude rewards. We show that this new visual deadly triad causes unstable training and premature convergence to degenerate solutions, a phenomenon we name catastrophic self-over…
DRExplainer: Quantifiable Interpretability in Drug Response Prediction with Directed Graph Convolutional Network
University of Science and Technology of China（中国科学技术大学）；School of Information and Artificial Intelligence, Anhui Agricultural University（安徽农业大学信息与人工智能学院）；Institutes of Physical Science and Information Technology, Anhui University（安徽大学物理科学与信息技术研究院）
Coflow is a recently proposed networking abstraction to help improve the communication performance of data-parallel computing jobs. In multi-stage jobs, each job consists of multiple coflows and is represented by a Directed Acyclic Graph (DAG). Efficiently scheduling coflows is critical to improve the data-parallel computing performance in data centers. Compared with hand-tuned scheduling heuristics, existing work DeepWeave [1] utilizes Reinforcement Learning (RL) framework to generate highly-efficient coflow scheduling policies automatically. It employs a graph neural network (GNN) to encode …
University of Lugano（卢加诺大学）；Harvard University（哈佛大学）；Tel Aviv University（特拉维夫大学）
Deep learning on graphs and in particular, graph convolutional neural networks, have recently attracted significant attention in the machine learning community. Many of such techniques explore the analogy between the graph Laplacian eigenvectors and the classical Fourier basis, allowing to formulate the convolution as a multiplication in the spectral domain. One of the key drawback of spectral CNNs is their explicit assumption of an undirected graph, leading to a symmetric Laplacian matrix with orthogonal eigendecomposition. In this work we propose MotifNet, a graph CNN capable of dealing with…
We propose a novel framework for value function factorization in multi-agent deep reinforcement learning (MARL) using graph neural networks (GNNs). In particular, we consider the team of agents as the set of nodes of a complete directed graph, whose edge weights are governed by an attention mechanism. Building upon this underlying graph, we introduce a mixing GNN module, which is responsible for i) factorizing the team state-action value function into individual per-agent observation-action value functions, and ii) explicit credit assignment to each agent in terms of fractions of the global te…
University of Washington（华盛顿大学）；Microsoft Research（微软研究院）；Georgia Institute of Technology（佐治亚理工学院）；NVIDIA（英伟达）
We consider the problem of learning motion policies for acceleration-based robotics systems with a structured policy class specified by RMPflow. RMPflow is a multi-task control framework that has been successfully applied in many robotics problems. Using RMPflow as a structured policy class in learning has several benefits, such as sufficient expressiveness, the flexibility to inject different levels of prior knowledge as well as the ability to transfer policies between robots. However, implementing a system for end-to-end learning RMPflow policies faces several computational challenges. In th…
