从特质竞争到贝叶斯推断
A Bird's-Eye View of Naming Game Dynamics: From Trait Competition to Bayesian Inference
摘要：
本贡献审查了一组不同版本的基本命名游戏模型，不同的基础拓扑结构或在调节代理之间的相互作用的机制。我们还包括最近推出的贝叶斯命名游戏模型，它融合了社会动态的基本命名游戏模型与贝叶斯学习框架介绍Tenenbaum和同事。后一种模型超越了标准符号动力学模型的名称和概念的固定性质以及相应的一次性学习过程，通过动态地描述代理如何根据贝叶斯推断的原理从几个示例中概括概念。
英文摘要：
The present contribution reviews a set of different versions of the basic naming game model, differing in the underlying topology or in the mechanisms regulating the interactions between agents. We include also a Bayesian naming game model recently introduced, which merges the social dynamics of the basic naming game model with the Bayesian learning framework introduced by Tenenbaum and co-workers. The latter model goes beyond the fixed nature of names and concepts of standard semiotic dynamics models and the corresponding one-shot learning process, by describing dynamically how agents can generalize a concept from a few examples, according to principles of Bayesian inference.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
