General policies represent reactive strategies for solving large families of planning problems like the infinite collection of solvable instances from a given domain. Methods for learning such policies from a collection of small training instances have been developed successfully for classical domains. In this work, we extend the formulations and the resulting combinatorial methods for learning general policies over fully observable, non-deterministic (FOND) domains. We also evaluate the resulting approach experimentally over a number of benchmark domains in FOND planning, present the general policies that result in some of these domains, and prove their correctness. The method for learning general policies for FOND planning can actually be seen as an alternative FOND planning method that searches for solutions, not in the given state space but in an abstract space defined by features that must be learned as well.
Korea Advanced Institute of Science and Technology (KAIST)（韩国科学技术院）；Pohang University of Science and Technology (POSTECH)（浦项科技大学）
Recently, graph-based planning algorithms have gained much attention to solve goal-conditioned reinforcement learning (RL) tasks: they provide a sequence of subgoals to reach the target-goal, and the agents learn to execute subgoal-conditioned policies. However, the sample-efficiency of such RL schemes still remains a challenge, particularly for long-horizon tasks. To address this issue, we present a simple yet effective self-imitation scheme which distills a subgoal-conditioned policy into the target-goal-conditioned policy. Our intuition here is that to reach a target-goal, an agent should p…
University of Massachusetts（马萨诸塞大学）；ICREA；Universitat Pompeu Fabra（庞培法布拉大学）
Fully observable non-deterministic (FOND) planning is becoming increasingly important as an approach for computing proper policies in probabilistic planning, extended temporal plans in LTL planning, and general plans in generalized planning. In this work, we introduce a SAT encoding for FOND planning that is compact and can produce compact strong cyclic policies. Simple variations of the encodings are also introduced for strong planning and for what we call, dual FOND planning, where some non-deterministic actions are assumed to be fair (e.g., probabilistic) and others unfair (e.g., adversaria…
We consider a settings of hierarchical reinforcement learning, in which the reward is a sum of components. For each component we are given a policy that maximizes it and our goal is to assemble a policy from the individual policies that maximizes the sum of the components. We provide theoretical guarantees for assembling such policies in deterministic MDPs with collectible rewards. Our approach builds on formulating this problem as a traveling salesman problem with discounted reward. We focus on local solutions, i.e., policies that only use information from the current state; thus, they are ea…
Australian National University（澳大利亚国立大学）；Data61, CSIRO（联邦科学与工业研究组织 Data61）；Data to Decisions CRC（数据决策合作研究中心）
In this paper, we introduce the Action Schema Network (ASNet): a neural network architecture for learning generalised policies for probabilistic planning problems. By mimicking the relational structure of planning problems, ASNets are able to adopt a weight-sharing scheme which allows the network to be applied to any problem from a given planning domain. This allows the cost of training the network to be amortised over all problems in that domain. Further, we propose a training method which balances exploration and supervised training on small problems to produce a policy which remains robust …
Singapore Management University（新加坡管理大学）；University of Toronto（多伦多大学）
Recent advances in deep learning have enabled optimization of deep reactive policies (DRPs) for continuous MDP planning by encoding a parametric policy as a deep neural network and exploiting automatic differentiation in an end-to-end model-based gradient descent framework. This approach has proven effective for optimizing DRPs in nonlinear continuous MDPs, but it requires a large number of sampled trajectories to learn effectively and can suffer from high variance in solution quality. In this work, we revisit the overall model-based DRP objective and instead take a minorization-maximization p…
State-of-the-art efficient model-based Reinforcement Learning (RL) algorithms typically act by iteratively solving empirical models, i.e., by performing \emph{full-planning} on Markov Decision Processes (MDPs) built by the gathered experience. In this paper, we focus on model-based RL in the finite-state finite-horizon MDP setting and establish that exploring with \emph{greedy policies} -- act by \emph{1-step planning} -- can achieve tight minimax performance in terms of regret, $\tilde{\mathcal{O}}(\sqrt{HSAT})$. Thus, full-planning in model-based RL can be avoided altogether without any perf…
Planning methods can solve temporally extended sequential decision making problems by composing simple behaviors. However, planning requires suitable abstractions for the states and transitions, which typically need to be designed by hand. In contrast, model-free reinforcement learning (RL) can acquire behaviors from low-level inputs directly, but often struggles with temporally extended tasks. Can we utilize reinforcement learning to automatically form the abstractions needed for planning, thus obtaining the best of both approaches? We show that goal-conditioned policies learned with RL can b…
Domain-independent probabilistic planners input an MDP description in a factored representation language such as PPDDL or RDDL, and exploit the specifics of the representation for faster planning. Traditional algorithms operate on each problem instance independently, and good methods for transferring experience from policies of other instances of a domain to a new instance do not exist. Recently, researchers have begun exploring the use of deep reactive policies, trained via deep reinforcement learning (RL), for MDP planning domains. One advantage of deep reactive policies is that they are mor…
In hierarchical planning for Markov decision processes (MDPs), temporal abstraction allows planning with macro-actions that take place at different time scale in form of sequential composition. In this paper, we propose a novel approach to compositional reasoning and hierarchical planning for MDPs under temporal logic constraints. In addition to sequential composition, we introduce a composition of policies based on generalized logic composition: Given sub-policies for sub-tasks and a new task expressed as logic compositions of subtasks, a semi-optimal policy, which is optimal in planning with…
Sapienza University of Rome（罗马大学（罗马第一大学））；Federal University of Rio Grande do Sul（南里奥格兰德联邦大学）
Fully Observable Non-Deterministic (FOND) planning models uncertainty through actions with non-deterministic effects. Existing FOND planning algorithms are effective and employ a wide range of techniques. However, most of the existing algorithms are not robust for dealing with both non-determinism and task size. In this paper, we develop a novel iterative depth-first search algorithm that solves FOND planning tasks and produces strong cyclic policies. Our algorithm is explicitly designed for FOND planning, addressing more directly the non-deterministic aspect of FOND planning, and it also expl…
It has been observed that in many of the benchmark planning domains, atomic goals can be reached with a simple polynomial exploration procedure, called IW, that runs in time exponential in the problem width. Such problems have indeed a bounded width: a width that does not grow with the number of problem variables and is often no greater than two. Yet, while the notion of width has become part of the state-of-the-art planning algorithms like BFWS, there is still no good explanation for why so many benchmark domains have bounded width. In this work, we address this question by relating bounded w…
VAMPS: Visual and Motor Policies from Sampling-Based Planning
LAAS-CNRS, Université de Toulouse, CNRS（LAAS-CNRS，图卢兹大学，法国国家科学研究中心）；Machines in Motion Laboratory, New York University（纽约大学运动机器实验室）；Industrial Engineering Department, University of Trento（特伦托大学工业工程系）；Artificial and Natural Intelligence Toulouse Institute (ANITI)（图卢兹人工与自然智能研究所（ANITI））
Imperial College London（帝国理工学院）；Fondazione Bruno Kessler（布鲁诺·凯斯勒基金会）
Heuristic forward search is currently the dominant paradigm in classical planning. Forward search algorithms typically rely on a single, relatively simple variation of best-first search and remain fixed throughout the process of solving a planning problem. Existing work combining multiple search techniques usually aims at supporting best-first search with an additional exploratory mechanism, triggered using a handcrafted criterion. A notable exception is very recent work which combines various search techniques using a trainable policy. It is, however, confined to a discrete action space compr…
In this work, we provide theoretical guarantees for reward decomposition in deterministic MDPs. Reward decomposition is a special case of Hierarchical Reinforcement Learning, that allows one to learn many policies in parallel and combine them into a composite solution. Our approach builds on mapping this problem into a Reward Discounted Traveling Salesman Problem, and then deriving approximate solutions for it. In particular, we focus on approximate solutions that are local, i.e., solutions that only observe information about the current state. Local policies are easy to implement and do not r…
MIT Computer Science and Artificial Intelligence Laboratory（麻省理工学院计算机科学与人工智能实验室）
A longstanding objective in classical planning is to synthesize policies that generalize across multiple problems from the same domain. In this work, we study generalized policy search-based methods with a focus on the score function used to guide the search over policies. We demonstrate limitations of two score functions and propose a new approach that overcomes these limitations. The main idea behind our approach, Policy-Guided Planning for Generalized Policy Generation (PG3), is that a candidate policy should be used to guide planning on training problems as a mechanism for evaluating that …
Linköping University（林雪平大学）；Universitat Pompeu Fabra（庞培法布拉大学）；Institució Catalana de Recerca i Estudis Avançats（加泰罗尼亚研究与高等教育学院）
We consider the problem of learning generalized policies for classical planning domains using graph neural networks from small instances represented in lifted STRIPS. The problem has been considered before but the proposed neural architectures are complex and the results are often mixed. In this work, we use a simple and general GNN architecture and aim at obtaining crisp experimental results and a deeper understanding: either the policy greedy in the learned value function achieves close to 100% generalization over instances larger than those used in training, or the failure must be understoo…
Discovering Effective Policies for Land-Use Planning with Neuroevolution
Cognizant AI Labs；The University of Texas at Austin（德克萨斯大学奥斯汀分校）；Ludwig-Maximilians-Universität（路德维希-马克西米利安大学）；Open Climate Fix（开放气候修复）；Cognizant Technology Solutions
Grupo Oesía（奥西亚集团）；Universitat Pompeu Fabra（庞培法布拉大学）
Optimal action selection in decision problems characterized by sparse, delayed rewards is still an open challenge. For these problems, current deep reinforcement learning methods require enormous amounts of data to learn controllers that reach human-level performance. In this work, we propose a method that interleaves planning and learning to address this issue. The planning step hinges on the Iterated-Width (IW) planner, a state of the art planner that makes explicit use of the state representation to perform structured exploration. IW is able to scale up to problems independently of the size…
Offline learning is a key part of making reinforcement learning (RL) useable in real systems. Offline RL looks at scenarios where there is data from a system's operation, but no direct access to the system when learning a policy. Recent work on training RL policies from offline data has shown results both with model-free policies learned directly from the data, or with planning on top of learnt models of the data. Model-free policies tend to be more performant, but are more opaque, harder to command externally, and less easy to integrate into larger systems. We propose an offline learner that …
Generalized planning is concerned with the computation of general policies that solve multiple instances of a planning domain all at once. It has been recently shown that these policies can be computed in two steps: first, a suitable abstraction in the form of a qualitative numerical planning problem (QNP) is learned from sample plans, then the general policies are obtained from the learned QNP using a planner. In this work, we introduce an alternative approach for computing more expressive general policies which does not require sample plans or a QNP planner. The new formulation is very simpl…
Google AI（谷歌人工智能）；University of New Mexico（新墨西哥大学）
This paper addresses two challenges facing sampling-based kinodynamic motion planning: a way to identify good candidate states for local transitions and the subsequent computationally intractable steering between these candidate states. Through the combination of sampling-based planning, a Rapidly Exploring Randomized Tree (RRT) and an efficient kinodynamic motion planner through machine learning, we propose an efficient solution to long-range planning for kinodynamic motion planning. First, we use deep reinforcement learning to learn an obstacle-avoiding policy that maps a robot's sensor obse…
This paper builds upon recent work in the declarative design of dialogue agents and proposes an exciting new tool -- D3BA -- Declarative Design for Digital Business Automation, built to optimize business processes using the power of AI planning. The tool provides a powerful framework to build, optimize, and maintain complex business processes and optimize them by composing with services that automate one or more subtasks. We illustrate salient features of this composition technique, compare with other philosophies of composition, and highlight exciting opportunities for research in this emergi…
Columbia University（哥伦比亚大学）；Massachusetts Institute of Technology（麻省理工学院）
The problem of retrosynthetic planning can be framed as one player game, in which the chemist (or a computer program) works backwards from a molecular target to simpler starting materials though a series of choices regarding which reactions to perform. This game is challenging as the combinatorial space of possible choices is astronomical, and the value of each choice remains uncertain until the synthesis plan is completed and its cost evaluated. Here, we address this problem using deep reinforcement learning to identify policies that make (near) optimal reaction choices during each step of re…
This paper addresses the problem of planning a safe (i.e., collision-free) trajectory from an initial state to a goal region when the obstacle space is a-priori unknown and is incrementally revealed online, e.g., through line-of-sight perception. Despite its ubiquitous nature, this formulation of motion planning has received relatively little theoretical investigation, as opposed to the setup where the environment is assumed known. A fundamental challenge is that, unlike motion planning with known obstacles, it is not even clear what an optimal policy to strive for is. Our contribution is thre…
This paper addresses the problem of reliably and efficiently solving broad classes of long-horizon stochastic path planning problems. Starting with a vanilla RL formulation with a stochastic dynamics simulator and an occupancy matrix of the environment, our approach computes useful options with policies as well as high-level paths that compose the discovered options. Our main contributions are (1) data-driven methods for creating abstract states that serve as endpoints for helpful options, (2) methods for computing option policies using auto-generated option guides in the form of dense pseudo-…
Yale University（耶鲁大学）；Northwestern University（西北大学）；Mercari, Inc.（Mercari 株式会社）；University of Wisconsin-Madison（威斯康星大学麦迪逊分校）
Off-policy evaluation (OPE) attempts to predict the performance of counterfactual policies using log data from a different policy. We extend its applicability by developing an OPE method for a class of both full support and deficient support logging policies in contextual-bandit settings. This class includes deterministic bandit (such as Upper Confidence Bound) as well as deterministic decision-making based on supervised and unsupervised learning. We prove that our method's prediction converges in probability to the true performance of a counterfactual policy as the sample size increases. We v…
Decision making in multi-agent systems (MAS) is a great challenge due to enormous state and joint action spaces as well as uncertainty, making centralized control generally infeasible. Decentralized control offers better scalability and robustness but requires mechanisms to coordinate on joint tasks and to avoid conflicts. Common approaches to learn decentralized policies for cooperative MAS suffer from non-stationarity and lacking credit assignment, which can lead to unstable and uncoordinated behavior in complex environments. In this paper, we propose Strong Emergent Policy approximation (ST…
Multi-Criteria Optimal Planning for Energy Policies in CLP
Università di Ferrara（费拉拉大学）；University of Bristol（布里斯托尔大学）；Università di Bologna（博洛尼亚大学）；Free University of Bozen-Bolzano（博尔扎诺自由大学）；ARPA Emilia-Romagna（艾米利亚-罗马涅大区环境保护局）
In this paper we present the first safe system for full control of self-driving vehicles trained from human demonstrations and deployed in challenging, real-world, urban environments. Current industry-standard solutions use rule-based systems for planning. Although they perform reasonably well in common scenarios, the engineering complexity renders this approach incompatible with human-level performance. On the other hand, the performance of machine-learned (ML) planning solutions can be improved by simply adding more exemplar data. However, ML methods cannot offer safety guarantees and someti…
The synthesis problem for partially observable Markov decision processes (POMDPs) is to compute a policy that satisfies a given specification. Such policies have to take the full execution history of a POMDP into account, rendering the problem undecidable in general. A common approach is to use a limited amount of memory and randomize over potential choices. Yet, this problem is still NP-hard and often computationally intractable in practice. A restricted problem is to use neither history nor randomization, yielding policies that are called stationary and deterministic. Previous approaches to …
In order to solve complex, long-horizon tasks, intelligent robots need to carry out high-level, abstract planning and reasoning in conjunction with motion planning. However, abstract models are typically lossy and plans or policies computed using them can be unexecutable. These problems are exacerbated in stochastic situations where the robot needs to reason about, and plan for multiple contingencies. We present a new approach for integrated task and motion planning in stochastic settings. In contrast to prior work in this direction, we show that our approach can effectively compute integrated…
Paderborn University（帕德博恩大学）；Karlstad University（卡尔斯塔德大学）；Hasso-Plattner-Institute（哈索·普拉特纳研究所）；University of Potsdam（波茨坦大学）
We present Distributed Deep Deterministic Policy Gradient (3DPG), a multi-agent actor-critic (MAAC) algorithm for Markov games. Unlike previous MAAC algorithms, 3DPG is fully distributed during both training and deployment. 3DPG agents calculate local policy gradients based on the most recently available local data (states, actions) and local policies of other agents. During training, this information is exchanged using a potentially lossy and delaying communication network. The network therefore induces Age of Information (AoI) for data and policies. We prove the asymptotic convergence of 3DP…
In order to solve complex, long-horizon tasks, intelligent robots need to carry out high-level, abstract planning and reasoning in conjunction with motion planning. However, abstract models are typically lossy and plans or policies computed using them can be inexecutable. These problems are exacerbated in stochastic situations where the robot needs to reason about and plan for multiple contingencies. We present a new approach for integrated task and motion planning in stochastic settings. In contrast to prior work in this direction, we show that our approach can effectively compute integrated …
Grupo Oesía（奥西亚集团）；Universitat Pompeu Fabra（庞培法布拉大学）
Width-based planning has demonstrated great success in recent years due to its ability to scale independently of the size of the state space. For example, Bandres et al. (2018) introduced a rollout version of the Iterated Width algorithm whose performance compares well with humans and learning methods in the pixel setting of the Atari games suite. In this setting, planning is done on-line using the "screen" states and selecting actions by looking ahead into the future. However, this algorithm is purely exploratory and does not leverage past reward information. Furthermore, it requires the stat…
