Kemmerich, Thomas: Learning and coordination in sequential multiagent problems. 2012
Inhalt
- Introduction
- Part I General Background
- Background and Related Work
- Multiagent Systems
- The Online Partitioning Problem
- Multirobot Task Allocation
- Swarm Intelligence
- Transfer Learning
- Optimization
- Reinforcement Learning
- Part II Foundations
- Sequential Stage Games and Engineered Rewards
- The Iterative Agent Partitioning Problem
- Storage Media
- Part III Algorithms
- Reinforcement Learning with Noisy Rewards
- Basic Definitions
- Deterministic Q-Learning with Noisy Rewards
- Q-Learning with Noisy Rewards and Optimistic Updates
- Conclusion
- Distributed Stateless Learning
- Introduction
- Approach
- Formal Analysis
- Evaluation
- Complexity
- Distributed Stateless Learning with Noisy Rewards
- Discussion
- Coordination in Sequential Stage Games
- Empirical Analysis
- Settings
- Basic Distributed Stateless Learning
- Strategy Specific Simulations
- Reward Types and Coordination Strategies
- Evaluation in Different Scenarios
- Discussion
- Storage Medium-Based Distributed Stateless Learning
- Region-Based Heuristics
- Part IV Conclusion
- Part V Appendix