We introduce a new algorithm, Regression based Supervised Learning (RSL), for learning per instance Neural Network (NN) defined heuristic functions for classical planning problems. RSL uses regression to select relevant sets of states at a range of different distances from the goal. RSL then formulates a Supervised Learning problem to obtain the parameters that define the NN heuristic, using the selected states labeled with exact or estimated distances to goal states. Our experimental study shows that RSL outperforms, in terms of coverage, previous classical planning NN heuristics functions while requiring a fraction of the training time.

We propose new width-based planning and learning algorithms inspired from a careful analysis of the design decisions made by previous width-based planners. The algorithms are applied over the Atari-2600 games and our best performing algorithm, Novelty guided Critical Path Learning (N-CPL), outperforms the previously introduced width-based planning and learning algorithms $\pi$-IW(1), $\pi$-IW(1)+ and $\pi$-HIW(n, 1). Furthermore, we present a taxonomy of the Atari-2600 games according to some of their defining characteristics. This analysis of the games provides further insight into the behaviour and performance of the algorithms introduced. Namely, for games with large branching factors, and games with sparse meaningful rewards, N-CPL outperforms $\pi$-IW, $\pi$-IW(1)+ and $\pi$-HIW(n, 1).

The automatic generation of realistic behavior such as tactical intercepts for Unmanned Aerial Vehicles (UAV) in air combat is a challenging problem. State-of-the-art solutions propose hand-crafted algorithms and heuristics whose performance depends heavily on the initial conditions and specific aerodynamic characteristics of the UAVs involved. This demo shows the ability of domain-independent planners, embedded into simulators, to generate on-line, feed-forward, control signals that steer simulated aircraft as best suits the situation.

Proper epistemic knowledge bases (PEKBs) are syntactic knowledge bases that use multi-agent epistemic logic to represent nested multi-agent knowledge and belief. PEKBs have certain syntactic restrictions that lead to desirable computational properties; primarily, a PEKB is a conjunction of modal literals, and therefore contains no disjunction. Sound entailment can be checked in polynomial time, and is complete for a large set of arbitrary formulae in logics Kn and KDn. In this paper, we extend PEKBs to deal with a restricted form of disjunction: 'knowing whether.' An agent i knows whether Q iff agent i knows Q or knows not Q; that is, []Q or []not(Q). In our experience, the ability to represent that an agent knows whether something holds is useful in many multi-agent domains. We represent knowing whether with a modal operator, and present sound polynomial-time entailment algorithms on PEKBs with the knowing whether operator in Kn and KDn, but which are complete for a smaller class of queries than standard PEKBs.

We present a situation calculus-based account of multi-players synchronous games in the style of general game playing. Such games can be represented as action theories of a special form, situation calculus synchronous game structures (SCSGSs), in which we have a single action tick whose effects depend on the combination of moves selected by the players. Then one can express properties of the game, e.g., winning conditions, playability, weak and strong winnability, etc. in a first-order alternating-time μ-calculus. We discuss verification in this framework considering computational effectiveness. We also show that SCSGSs can be considered as a first-order variant of the Game Description Language (GDL) that supports infinite domains and possibly non-terminating games. We do so by giving a translation of GDL specifications into SCSGSs and showing its correctness. Finally, we show how a player's possible moves can be specified in a Golog-like programming language.

Many AI applications involve the interaction of multiple autonomous agents, requiring those agents to reason about their own beliefs, as well as those of other agents. However, planning involving nested beliefs is known to be computationally challenging. In this work, we address the task of synthesizing plans that necessitate reasoning about the beliefs of other agents. We plan from the perspective of a single agent with the potential for goals and actions that involve nested beliefs, non-homogeneous agents, co-present observations, and the ability for one agent to reason as if it were another. We formally characterize our notion of planning with nested belief, and subsequently demonstrate how to automatically convert such problems into problems that appeal to classical planning technology. Our approach represents an important first step towards applying the well-established field of automated planning to the challenging task of planning involving nested beliefs of multiple agents.

We consider the short-term production scheduling problem for a network of multiple open-pit mines and ports. Ore produced at each mine is transported by rail to a set of ports and blended into signature products for shipping. Consistency in the grade and quality of production over time is critical for customer satisfaction, whereas the maximal production of blended products is required to maximise profit. In practice, short-term schedules are formed independently at each mine, tasked with achieving the grade and quality targets outlined in a medium-term plan. However, because of uncertainty in the data available to a medium-term planner and the dynamics of the mining environment, such targets may not be feasible in the short term. We present a decomposition-based heuristic for this short-term scheduling problem in which the grade and quality goals assigned to each mine are collaboratively adapted—ensuring the satisfaction of blending constraints at each port and exploiting opportunities to maximise production in the network that would otherwise be missed.

A new learning algorithm, the Consolidated Learning Algorithm based on Relational Evidence Theory (CLARET) is presented, which relies on inductive logic programming and graph matching techniques. In relational evidence theory, two different approaches to evidential learning are consolidated in how they apply to generalising within relational data structures. Attribute-based discrimination (decision trees) is integrated with part-based interpretation (graph matching) for evaluating and updating representations in spatial domains. This allows an interpretation stage to be incorporated into the generalization process. A systematic approach to finding interpretations is demonstrated for an online, hand drawn, schematic diagram and symbols recognition system. The approach uses an adaptive representational bias and search strategy during learning by efficiently grounding the learning procedures in the relational spatial constraints of their application.