基于消息传播的Bayesian推理算法研究

Researching on Message-based Transmission Bayesian Inference Algorithm

利用网络结构特性和条件独立性之间的关系设计推理有效方法,是Bayesian网络推理的主要研究内容.各种推理算法的区别主要在于计算速度.在深入讨论已有算法的基础上,设计并实现了基于消息传播的Bayesian推理算法,有效地降低了推理过程中运算的时空复杂度.

It is the main research content of Bayesian network to design effective way of reasoning using structural characteristics and independence conditions.The major difference between the various inference algorithms is the calculation speed.In this paper,on discussing in depth the existing algorithms we designed and realized a messagebased transmission Bayesian inference algorithm,which reduced the reasoning process by time and space complexity.