Between MDPs and semi-MDPs: a framework for temporal abstraction in reinforcement learning
Artificial Intelligence
Nearly deterministic abstractions of Markov decision processes
Eighteenth national conference on Artificial intelligence
Planning, learning and coordination in multiagent decision processes
TARK '96 Proceedings of the 6th conference on Theoretical aspects of rationality and knowledge
Hierarchical solution of Markov decision processes using macro-actions
UAI'98 Proceedings of the Fourteenth conference on Uncertainty in artificial intelligence
Editorial: The traveling salesman problem
Discrete Optimization
Design, runtime, and analysis of multi-agent systems
Proceedings of the fourth international joint conference on Autonomous agents and multiagent systems
DAMAS'05 Proceedings of the 2005 international conference on Defence Applications of Multi-Agent Systems
Agent technology for coordinating UAV target tracking
KES'05 Proceedings of the 9th international conference on Knowledge-Based Intelligent Information and Engineering Systems - Volume Part II
Simulating UAV Surveillance for Analyzing Impact of Commitments in Multi-Agent Systems
International Journal of Agent Technologies and Systems
Hi-index | 0.00 |
Autonomous agents are given the authority to select which actions they will execute. If the agent behaves rationally, the actions it selects will be in its own best interests. When addressing multiple goals, the rational action may not be obvious. Equipping the agents with decision-theoretic methods allows the agent to mathematically evaluate the risks, uncertainty, and benefits of the various available courses of action. Using this evaluation, an agent can determine which goals are worth achieving, as well as the order in which to achieve those goals. When the goals of the agent changes, the agent must replan to maintain rational decision-making. This research uses macro actions to transform the state space for the agent's decision problem into the desire space of the agent. Reasoning in the desire space, the agent can efficiently maintain rationality in response to addition and removal of goals.