Between MDPs and semi-MDPs: a framework for temporal abstraction in reinforcement learning
Artificial Intelligence
Modeling task allocation using a decision theoretic model
Proceedings of the fourth international joint conference on Autonomous agents and multiagent systems
Hierarchical multi-agent reinforcement learning
Autonomous Agents and Multi-Agent Systems
Hierarchical reinforcement learning with the MAXQ value function decomposition
Journal of Artificial Intelligence Research
The communicative multiagent team decision problem: analyzing teamwork theories and models
Journal of Artificial Intelligence Research
Sequential optimality and coordination in multiagent systems
IJCAI'99 Proceedings of the 16th international joint conference on Artifical intelligence - Volume 1
The multi-team formation precursor of teamwork
EPIA'05 Proceedings of the 12th Portuguese conference on Progress in Artificial Intelligence
A hybrid approach to multi-agent decision-making
Proceedings of the 2008 conference on ECAI 2008: 18th European Conference on Artificial Intelligence
Hi-index | 0.00 |
The response to a large-scale disaster, e.g. an earthquake or a terrorist incident, urges for low-cost policies that coordinate sequential decisions of multiple agents. Decisions range from collective (common good) to individual (self-interested) perspectives, intuitively shaping a two-layer decision model. However, current decision theoretic models are either purely collective or purely individual and seek optimal policies. We present a two-layer, collective versus individual (CvI) decision model and explore the tradeoff between cost reduction and loss of optimality while learning coordination skills. Experiments, in a partially observable domain, test our approach for learning a collective policy and results show near-optimal policies that exhibit coordinated behavior.