论文部分内容阅读
给出了如何利用已知数据集,在因果图结构已知的条件下学习或估计出连接强度的方法,包括完备数据情形。采用贝叶斯方法,利用后验分布的数学期望学习因果图连接强度的条件期望估计法和不完备数据的类似期望最大化(EM)学习因果图连接强度的算法。并用实例验证了算法的有效性和可行性。
How to use the known data set to learn or estimate the connection strength under the condition of the causal structure is given, including the complete data case. Using Bayesian method, the mathematical expectation of posterior distribution is used to study the similarity expectation maximization (EM) of conditional expectation estimation and incomplete data of causality graph connection strength. An example is given to verify the effectiveness and feasibility of the algorithm.