Bayes-optimal behavior, while well-defined, is often difficult to achieve. Recent advances in the use of Monte-Carlo tree search (MCTS) have shown that it is possible to act near-optimally in Markov Decision Processes (MDPs) with very large or infinite state spaces. Bayes-optimal behavior in an unknown MDP is equivalent to optimal behavior in the known belief-space MDP, although the size of this belief-space MDP grows exponentially with the amount of history retained, and is potentially infinite. We show how an agent can use one particular MCTS algorithm, Forward Search Sparse Sampling (FSSS), in an efficient way to act nearly Bayes-optimally for all but a polynomial number of steps, assuming that FSSS can be used to act efficiently in any possible underlying MDP.

This paper presents an approach for learning to translate simple narratives, i.e., texts (sequences of sentences) describing dynamic systems, into coherent sequences of events without the need for labeled training data. Our approach incorporates domain knowledge in the form of preconditions and effects of events, and we show that it outperforms state-of-the-art supervised learning systems on the task of reconstructing RoboCup soccer games from their commentaries.

Scheduling problems in manufacturing, logistics and project management have frequently been modeled using the framework of Resource Constrained Project Scheduling Problems with minimum and maximum time lags (RCPSP/max). Due to the importance of these problems, providing scalable solution schedules for RCPSP/max problems is a topic of extensive research. However, all existing methods for solving RCPSP/max assume that durations of activities are known with certainty, an assumption that does not hold in real world scheduling problems where unexpected external events such as manpower availability, weather changes, etc. lead to delays or advances in completion of activities. Thus, in this paper, our focus is on providing a scalable method for solving RCPSP/max problems with durational uncertainty. To that end, we introduce the robust local search method consisting of three key ideas: (a) Introducing and studying the properties of two decision rule approximations used to compute start times of activities with respect to dynamic realizations of the durational uncertainty; (b) Deriving the expression for robust makespan of an execution strategy based on decision rule approximations; and (c) A robust local search mechanism to efficiently compute activity execution strategies that are robust against durational uncertainty. Furthermore, we also provide enhancements to local search that exploit temporal dependencies between activities. Our experimental results illustrate that robust local search is able to provide robust execution strategies efficiently.

We propose the concept of Action-Related Place (ARPlace) as a powerful and flexible representation of task-related place in the context of mobile manipulation. ARPlace represents robot base locations not as a single position, but rather as a collection of positions, each with an associated probability that the manipulation action will succeed when located there. ARPlaces are generated using a predictive model that is acquired through experience-based learning, and take into account the uncertainty the robot has about its own location and the location of the object to be manipulated. When executing the task, rather than choosing one specific goal position based only on the initial knowledge about the task context, the robot instantiates an ARPlace, and bases its decisions on this ARPlace, which is updated as new information about the task becomes available. To show the advantages of this least-commitment approach, we present a transformational planner that reasons about ARPlaces in order to optimize symbolic plans. Our empirical evaluation demonstrates that using ARPlaces leads to more robust and efficient mobile manipulation in the face of state estimation uncertainty on our simulated robot.

This paper introduces two new frameworks for learning action models for planning. In the mistake-bounded planning framework, the learner has access to a planner for the given model representation, a simulator, and a planning problem generator, and aims to learn a model with at most a polynomial number of faulty plans. In the planned exploration framework, the learner does not have access to a problem generator and must instead design its own problems, plan for them, and converge with at most a polynomial number of planning attempts. The paper reduces learning in these frameworks to concept learning with one-sided error and provides algorithms for successful learning in both frameworks. A specific family of hypothesis spaces is shown to be efficiently learnable in both the frameworks.

Monte-Carlo Tree Search (MCTS) has proven to be a powerful, generic planning technique for decision-making in single-agent and adversarial environments. The stochastic nature of the Monte-Carlo simulations introduces errors in the value estimates, both in terms of bias and variance. Whilst reducing bias (typically through the addition of domain knowledge) has been studied in the MCTS literature, comparatively little effort has focused on reducing variance. This is somewhat surprising, since variance reduction techniques are a well-studied area in classical statistics. In this paper, we examine the application of some standard techniques for variance reduction in MCTS, including common random numbers, antithetic variates and control variates. We demonstrate how these techniques can be applied to MCTS and explore their efficacy on three different stochastic, single-agent settings: Pig, Can't Stop and Dominion.

Hierarchical Task Network (HTN) planning uses task decomposition to plan for an executable sequence of actions as a solution to a problem. In order to reason effectively, an HTN planner needs expressive domain knowledge. For instance, a simplified HTN planning system such as JSHOP2 uses such expressivity and avoids some task interactions due to the increased complexity of the planning process. We address the possibility of simplifying the domain representation needed for an HTN planner to find good solutions, especially in real-world domains describing home and building automation environments. We extend the JSHOP2 planner to reason about task interaction that happens when task's effects are already achieved by other tasks. The planner then prunes some of the redundant searches that can occur due to the planning process's interleaving nature. We evaluate the original and our improved planner on two benchmark domains. We show that our planner behaves better by using simplified domain knowledge and outperforms JSHOP2 in a number of relevant cases.

The task of keyhole (unobtrusive) plan recognition is central to adaptive game AI. "Tech trees" or "build trees" are the core of real-time strategy (RTS) game strategic (long term) planning. This paper presents a generic and simple Bayesian model for RTS build tree prediction from noisy observations, which parameters are learned from replays (game logs). This unsupervised machine learning approach involves minimal work for the game developers as it leverage players' data (com- mon in RTS). We applied it to StarCraft1 and showed that it yields high quality and robust predictions, that can feed an adaptive AI.

This paper explores six different representations of the BlocksWorld Domain. It compares the results of seven planners run on these representations. It shows that the rankings for the International Planning Competition, using the non-satisficing scoring function, would change for every representation.

An example is a Mars rover: The major problem with applying POMDP approaches to thanks to low-level control and obstacle avoidance, rovers realistic planning problems like the Mars rovers is the sheer can be expected to reach their destinations reliably, and can size of the problems. Using point-based approximations and collect and communicate data, but they do not know in advance structured representations similar to those used in classical which science targets are interesting and hence will planning (Poupart 2005), problems with tens of millions provide valuable data. Similarly, robots performing tasks of states can be solved approximately, but even that corresponds such as security or cognitive assistance are generally able to to a classical planning problem with only 25 binary navigate reliably, but use unreliable vision algorithms to detect variables, which is a quite small problem by the standards the people and objects with which they are supposed of classical deterministic planning. The alternative we propose to interact. Following Besse and Chaib-draa (2009), we in this paper is to construct a series of classical deterministic will refer to problems with deterministic actions but stochastic planning problems from the quasi-deterministic observations as quasi-deterministic problems, which differ problem. By solving each of these deterministic problems from Deterministic-POMDPs (DET-POMDPS) (Bonet we construct a contingent plan--one that contains branches 2009) by taking into account of uncertainty from observation to be chosen between at run-time.