多值动态不确定因果图的参数学习研究  

Research on Parameter Learning of Multi-Value Dynamic Uncertain Causality Graph

作　　者：何大材 王洪春 HE Dacai;WANG Hongcun(School of Mathematical Sciences,Chongqing Normal University,Chongqing 401331,China)

机构地区：[1]重庆师范大学数学科学学院,重庆401331

出　　处：《东莞理工学院学报》2025年第1期19-25,共7页Journal of Dongguan University of Technology

基　　金：国家社会科学基金一般项目(13BTJ008);重庆市教育委员会人文社会科学重点项目(22SKGH081);重庆市教育委员会重庆市高等教育教学改革研究项目(213139)。

摘　　要：收集一个复杂系统的多值动态不确定因果图的全局数据是不现实的,但是收集其局部数据相对容易。现有运用局部数据求解多值动态不确定因果图参数的方法是利用模型本身的性质,将参数学习转变为带等式约束条件的非线性规划问题。在此基础上提出一种两步式求解参数的方法,该方法通过多值动态不确定因果图的结构特点及局部数据对应的因果图子图,将需要学习的参数分为两部分,依次求解出参数。该方法与原方法相比,在求解第一部分参数时,通过更充分利用局部数据,提高了求解参数的准确性;在求解第二部分参数时,通过减少约束条件,简化了数学模型,提高了计算效率。It is unrealistic to collect global data of a complex system's multi-value dynamic uncertain causality graph,but it is relatively easy to collect its local data.The existing method to solve the parameters of the multi-value dynamic uncertain causality graph using local data is to transform it into a nonlinear programming problem with equality constraints through the nature of the model itself,and then solve the optimal solution to obtain the parameters.Based on this,this paper propose a two-step method for solving the parameters.This method solves the parameters into two parts through the structural characteristics of the multi-value dynamic uncertain causal graph and the causal graph subgraph corresponding to the local data.The parameters in the first part can be directly obtained from the local data so that the local data can be better used and the parameters are more accurate;the learning of the parameters in the second part compared with the original method,the number of obtained parameters and the constraints are reduced,so the calculation difficulty is reduced and the calculation efficiency is improved.

关 键 词：多值动态不确定因果图 参数学习 最优解 

分 类 号：O157.5[理学—数学]