科学确证的贝叶斯网络模型  

Bayesian Network Model of Scientific Confirmation

作　　者：张顺[1] 

机构地区：[1]南京大学哲学系,南京210023

出　　处：《重庆理工大学学报（社会科学）》2016年第6期14-20,66,共8页Journal of Chongqing University of Technology（Social Science）

基　　金：国家社会科学基金项目"归纳悖论与确证逻辑新探"(11BZX061);江苏省社会科学基金项目"形式知识论研究"(10ZXC009)

摘　　要：科学确证的假说-演绎模型与贝叶斯确证模型分别体现了定性确证与定量确证的内涵,然而科学确证是一个在纵横方向都具备复杂内涵的问题,两种模型均没克服因确证本身的复杂性与非单调性所带来的一些难题。为此,将贝叶斯网络方法运用到确证理论的建构中,通过图论与贝叶斯确证理论的结合开拓确证理论研究的新进路,并构建贝叶斯确证网络的简单模型。同时,从模型的定义、结构建模、推理等方面阐述贝叶斯确证网络的基本内容。The two classical confirmation theories-hypothetico-deductivism and Bayesian confirmation model respectively embody the connotation of qualitative confirmation and quantitative confirmation.However,scientific confirmation is an issue with complex content in its all directions. The two models fail to conquer the complexity and nonmonotonicity of the theory itself. This paper attempts to apply the Bayesian network method to the construction of confirmation theory. The combination of graph theory and Bayesian confirmation theory opens up a new research pathway of confirmation theory.Meanwhile,this paper tries to construct simple models of Bayesian confirmation network and elaborates this network＇s basic content in terms of its definition,structure modeling,reasoning etc.

关 键 词：科学确证 贝叶斯确证 贝叶斯确证网络 

分 类 号：B81[哲学宗教—逻辑学]