基于粗集不相容系统的膨胀土分类规则提取

分析膨胀土分类的粗糙性, 指出膨胀土分类是一个基于粗糙集的信息不相容决策系统. 针对常规方法容易引起规则失真的不足, 提出将贝叶斯理论和不相容系统决策挖掘相结合来提取膨胀土分类规则： 以膨胀土分类决策系统的可信度为先验概率, 膨胀土试验数据的支持度为后验概率, 计算膨胀土分类规则的条件概率;提取条件概率大于某一阈值的规则;通过逻辑合取与析取归并膨胀土分类规则. 实例计算和应用分析结果表明： 采用贝叶斯理论和基于粗糙集的不相容系统决策挖掘相结合的方法有利于基于粗糙集的不相容系统的数据挖掘, 而且为膨胀土分类规则的提取提供了一种切实可行的算法.

Generating expansive soil classification rule based on rough sets for inconsistent decision system

Expansive soil classification is an inconsistent system based on rough sets after analyzing its roughness. To avoid rule distortion resulting from routine methods, expansive soil classification rule was generated by utilizing the Bayes theory and data mining based on rough sets for inconsistent decision system. The reliability of expansive soil classification decision system was regarded as the priori probability, and the support degree of expansive soil data from experiment was regarded as posterior probability, and then the conditional probability was calculated. The rules were preserved that conditional probability was bigger than a given threshold value. The rule of expansive soil classification was generated through logic conjunction and disjunction to all of the preserved rules. The results show that the method not only perfects data mining based on rough sets for inconsistent decision system, but also provides an efficient and feasible algorithm to classify expansive soil.