In this paper we outline a general approach to the study of metareasoning, not in the sense of explicating the semantics of explicitly specified meta-level control policies, but in the sense of providing a basis for selecting and justifying computational actions. This research contributes to a developing attack on the problem of resource-bounded rationality, by providing a means for analysing and generating optimal computational strategies. Because reasoning about a computation without doing it necessarily involves uncertainty as to its outcome, probability and decision theory will be our main tools. We develop a general formula for the utility of computations, this utility being derived directly from the ability of computations to affect an agent's external actions. We address some philosophical difficulties that arise in specifying this formula, given our assumption of limited rationality. We also describe a methodology for applying the theory to particular problem-solving systems, a...

Despite their different perspectives, artificial intelligence (AI) and the disciplines of decision science have common roots and strive for similar goals. This paper surveys the potential for addressing problems in representation, inference, knowledge engineering, and explanation within the decision-theoretic framework. Recent analyses of the restrictions of several traditional AI reasoning techniques, coupled with the development of more tractable and expressive decisiontheoretic representation and inference strategies, have stimulated renewed interest in decision theory and decision analysis. We describe early experience with simple probabilistic schemes for automated reasoning, review the dominant expert-system paradigm, and survey some recent research at the crossroads of AI and decision science. In particular, we present the belief network and influence diagram representations. Finally, we discuss issues that have not been studied in detail within the expert-systems sett...

We define and exercise the expected value of computation as a fundamental component of reflection about alternative inference strategies. We present a portion of Protos research focused on the interlacing of reflection and action under scarce resources, and discuss how the techniques have been applied in a high-stakes medical domain. The work centers on endowing a computational agent with the ability to harness incomplete characterizations of problemsolving performance to control the amount of effort applied to a problem or subproblem, before taking action in the world or turning to another problem. We explore the use of the techniques in controlling decision-theoretic inference itself, and pose the approach as a model of rationality under scarce resources. 1 Reflection and Flexibility Reflection about the course of problem solving and about the interleaving of problem solving and physical activity is a hallmark of intelligent behavior. Applying a portion of available reasoning resour...

The long-term goal of our field is the creation and understanding of intelligence. Productive research in AI, both practical and theoretical, benefits from a notion of intelligence that is precise enough to allow the cumulative development of robust systems and general results. This paper outlines a gradual evolution in our formal conception of intelligence that brings it closer to our informal conception and simultaneously reduces the gap between theory and practice. 1 Artificial Intelligence AI is a field in which the ultimate goal has often been somewhat ill-defined and subject to dispute. Some researchers aim to emulate human cognition, others aim at the creation of

We introduce and analyze the problem of the compilation of decision models from a decision-theoretic perspective. The techniques described allow us to evaluate various configurations of compiled knowledge given the nature of evidential relationships in a domain, the utilities associated with alternative actions, the costs of run-time delays, and the costs of memory. We describe procedures for selecting a subset of the total observations available to be incorporated into a compiled situation–action mapping, in the context of a binary decision with conditional independence of evidence. The methods allow us to incrementally select the best pieces of evidence to add to the set of compiled knowledge in an engineering setting. After presenting several approaches to compilation, we exercise one of the methods to provide insight into the relationship between the distribution over weights of evidence and the preferred degree of compilation. 1

In real world problems, time for deliberation is often limited. Anytime algorithms are beneficial in these conditions as they usually find a first, possibly highly suboptimal, solution very fast and then continually work on improving the solution until allocated time expires. While anytime algorithms are popular, existing anytime search methods are unable to provide a measure of goodness of their results. In this paper we propose the ARA* algorithm. ARA* is an anytime heuristic search which tunes its performance bound based on available search time. It starts by finding a suboptimal solution quickly using a loose bound, then tightens the bound progressively as time allows. Given enough time it finds a provably optimal solution. In addition to the theoretical analysis we demonstrate the practical utility of ARA* with experiments on a simulated robot kinematic arm and dynamic path planning problem for an outdoor rover.

We describe a family of recently developed heuristicbased algorithms used for path planning in the real world. We discuss the fundamental similarities between static algorithms (e.g. A*), replanning algorithms (e.g. D*), anytime algorithms (e.g. ARA*), and anytime replanning algorithms (e.g. AD*). We introduce the motivation behind each class of algorithms, discuss their use on real robotic systems, and highlight their practical benefits and disadvantages.

The ability to categorize and use concepts e#ectively is a basic requirementofany intelligent actor. The utility-based approach to categorization is founded on the thesis that categorization is fundamentally in service of action, i.e., the choice of concepts made by an actor is critical to its choice of appropriate actions. This is in contrast to classical and similarity-based approaches which seek logical completeness in concept description with respect to sensory data rather than action-oriented e#ectiveness. Utility-based categorization is normative and not descriptive. It prescribes howanintelligent agent ought to conceptualize to act e#ectively. It provides ideals for categorization, speci#es criteria for the design of e#ective computational agents, and provides a model of ideal competence. A decision-theoretic framework for utilitybased categorization whichinvolves reasoning about alternative categorization models of varying levels of abstraction is proposed. Categorization mode...

We present a direction of research on principles and methods that can enable general problem solving via human computation systems. A key challenge in human computation is the effective and efficient coordination of problem solving. While simple tasks may be easy to partition across individuals, more complex tasks highlight challenges and opportunities for more sophisticated coordination and optimization, leveraging such core notions as problem decomposition, subproblem routing and solution, and the recomposition of solved subproblems into solutions. We discuss the interplay between algorithmic paradigms and human abilities, and illustrate through examples how members of a crowd can play diverse roles in an organized problem-solving process, serving not only as ‘data oracles ’ at the endpoints of computation, but also as modules for decomposing problems, controlling the algorithmic progression, and performing human program synthesis.

Most planners constructed up to now are qualitative: they deal with uncertainty by considering all possible outcomes of each plan, without quantifying their relative likelihood. They then choose a plan that deals with the worstcase scenario. However, it is clearly infeasible to plan for every possible contingency. Even beyond the purely computational considerations, planning for highly unlikely worst-case scenarios can force the agent to choose an overly cautious plan with low utility. One common way to avoid this problem is to make assumptions about the behavior of the world, i.e., assume that certain contingencies are impossible. In this paper, we analyze the paradigm of qualitative planning under assumptions, using decision-theoretic tools. We present conditions that guarantee the existence of optimal assumptions (ones inducing the agent to choose the plan with maximum expected utility). Finally, we sketch how assumptions can be constructed for a certain restricted class of plannin...