Explaining and repairing plans that fail
Artificial Intelligence
Operational rationality through compilation of anytime algorithms
Operational rationality through compilation of anytime algorithms
Empirical methods for artificial intelligence
Empirical methods for artificial intelligence
Optimal composition of real-time systems
Artificial Intelligence
Computer science as empirical inquiry: symbols and search
Communications of the ACM
Managing emergent character-based narrative
Proceedings of the 2nd international conference on INtelligent TEchnologies for interactive enterTAINment
Knowledge-based anytime computation
IJCAI'95 Proceedings of the 14th international joint conference on Artificial intelligence - Volume 1
Plan-based control of robotic agents: improving the capabilities of autonomous robots
Plan-based control of robotic agents: improving the capabilities of autonomous robots
Scheduling with probability and temporal constraints
AI*IA'05 Proceedings of the 9th conference on Advances in Artificial Intelligence
Hi-index | 0.00 |
Evaluation and comparison of existing planning systems is hard because they disagree on the fundamental role of planning, on evaluation metrics, and on the notion of success and failure. This paper suggests a decision-theoretic approach to evaluate planning systems that generalizes the role of planning in intelligent systems. The planner is viewed as a source of information that is used by an execution architecture in order to select actions. A planner is only as good as the effect it has on the performance of an operational system. Our approach calls for a clear separation between the planning component and the execution architecture and for evaluation of planning systems within the context of a well-defined command, planning and execution environment. The evaluation is based on the expected utility of the domain histories that are induced by the planning component.