Probabilistic Graphical Models 2 Bayesian Network Fundamentals

来源:互联网 发布:绵阳长虹网络公司地址 编辑:程序博客网 时间:2024/04/23 22:13

================Semantics & Factorization============

1.第一个简单贝叶斯网络











==============Reasoning Patterns==============










============Flow of Probabilistic Influence=============





==============Conditional Independence=============

D,I间无连线,相互独立

=========Independencies in Bayesian Networks==========

以上理论其实都为下图服务,可以看到条件概率被彻底约间。利用(非父亲、非儿子理论)

===============Naive Bayes==============




================Application - Medical Diagnosis=============








================Knowledge Engineering Example==================








good student对应的no accident反而降低了。注意:good student-----》young升高,年轻人毛躁






注意AGE   BLOCKgood students 对 Driver_quality的影响





 

 

 

 






原创粉丝点击