arXiv · Artificial Intelligence· Jose Tupayachi, Xueping Li, Soham Das·· 7 小时前AI 评分20
Learning to Harvest Without Collapse in a Regenerative Commons: A Lagrangian Framework
Learning to Harvest Without Collapse in a Regenerative Commons: A Lagrangian Framework
AI 导读
论文提出非平稳拉格朗日框架，将可再生公共资源建模为带消耗预算约束的多智能体MDP，通过无约束博弈解构造约束策略序列。引入平均周期解概念与奖励无关的可行性证书，并在Gordon-Schaefer渔业实验中验证消耗预算对库存保留、收获奖励与价格适应的影响。
来源：arXiv · Artificial Intelligence · arxiv.org