认知学习与模拟  

On Cognitive Learning and Simulation

作　　者：叶风[1] 洪勇 王亚东[1] 徐晓飞[1] 

机构地区：[1]哈尔滨工业大学计算机科学与工程系应用软件教研室

出　　处：《哈尔滨工业大学学报》1997年第5期77-80,88,共5页Journal of Harbin Institute of Technology

摘　　要：认知学习模拟，特别是对顿悟的认知模拟，是人工智能尚未触及的困难领域之一。本文重点就顿悟认知模拟问题探讨其认知基础并提出一种模拟的计算模型；采用一种基于缺席逻辑（ＤＬ）的称作必要性假设的理论（ＮＬ）来描述这一模型，同时讨论与模型密切相关的背景知识理论与合理假设引入问题；最后以ＮＬ解决两个著名的认知学习实验问题以表明模拟的可能性。The simulation on cognitive learning, specially on insight cognitive, is one of the most difficult research fields that AI has not touched yet Focusing on the issue on the simulation related to insight cognitive, we study its cognitive foundation and present a computing model of simulation This model is then described with a theory called necessity hypothses logic (NL) which is based on default logic (DL) The issues on the background knowledge theory and the introduction of reasonable hypotheses are also discussed With NL we have solved two famous problems in cognitive learning experiment to illustrate the feasibility of simulation

关 键 词：认知学习 模拟 必要性假设 推理 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]