【摘要】 近年来在人工智能领域,不确定性问题一直成为人们关注和研究的焦点。贝叶斯网是用来表示不确定变量集合联合概率分布的图形模式,它反映了变量间潜在的依赖关系。使用贝叶斯网建模已成为解决许多不确定性问题的强有力工具。基于国内外最新的研究成果对贝叶斯网模型的学习、推理和应用情况进行了综述,并对未来的发展方向进行了展望。更多还原

【Abstract】 In recent years,uncertainty reasoning in Artificial Intelligence has been a focus of research.A Bayesian Belief Network(BBN)is a graphic model that encodes joint probability distribution among uncertain variables,it express a potential dependent relationship between variables.Modeling with Bayesian belief network has been a powerful tool to solve many uncertainty problems.Based on the latest researched results at home and abroad,this paper reviews the learning,inference and applications of the Bayesian Belief Network,and presents possible future research orientations.更多还原