论文部分内容阅读
近20年来,贝叶斯1网络(Bayesian network,BN)在人工智能领域受到了广泛关注,得到了深入研究,已有大量成熟的研究成果。贝叶斯网络是图论与概率论相结合的产物,直观地表示为一个复杂的赋值因果关系图,图中各节点表示所讨论的问题域中的变量或事件。节点之间的弧表示事件之间的直接因果关系。贝叶斯网络的实质就是所研究领域的概率分布。贝叶斯网络由于其坚实的数学理论基础,被认为是在不确定性环境中实现知识表示、推断、预测等最理想的工具。它已在数据挖掘、故障诊断、图像识别等领域得到了较好的应用。
In recent 20 years, Bayesian network (BN) has drawn much attention in the field of artificial intelligence and has been deeply studied. There have been a lot of mature research results. Bayesian network is a combination of graph theory and probability theory, which is intuitively represented as a complex causality diagram. Each node in the graph represents the variable or event in the domain in question. Arcs between nodes represent a direct causal relationship between events. The essence of Bayesian networks is the probability distribution of the area under study. Because of its solid mathematical theory, Bayesian network is considered as the most ideal tool to realize knowledge representation, inference and prediction in an uncertain environment. It has been in data mining, fault diagnosis, image recognition and other fields have been better applied.