罗捷婷:Strategic knowledge in multi-agent systems
Strategic knowledge in multi-agent systems
主讲人简介:
罗捷婷,浙江大学哲学学院百人计划研究员、博士生导师。2018年毕业于荷兰乌特勒支大学,获哲学博士学位。先后在多所国际知名高校和研究所从事博士后研究工作。主要研究领域为社会和认知现象(包括情感)的形式化和计算建模,并以人工智能为中心,结合社会学、心理学、经济学、哲学等开展交叉学科研究,以用于自主智能体系统的设计和开发。
报告简介:
“We introduce a logic of strategic knowledge where knowledge, including knowledge of strategies, is expressible in the object language via a novel predicate S_a(x) meaning agent a has strategy x. Unlike intermediate approaches that specify strategy knowledge outside the model, our semantics interprets knowledge solely through the indistinguishability relation of the concurrent game structure. We prove model-checking complexity using submodel preservation, and show that truncating a state's history preserves formula truth, a key step toward decidability of satisfiability.
