中文摘要：

在实际应用中,集合是动态变化的,包括元素的迁入和迁出.双向概率PS-粗糙集既考虑集合的动态特性,又考虑知识库中统计信息的粗集理论.在深入分析概率PS-粗糙集理论及其性质的基础上提出了基于双向概率PS-粗糙集的动态三支决策模型,以解决集合动态变化下的三支决策问题.首先,根据双向概率PS-粗糙集的上下近似得到概率PS-正域、负域和边界域,给出了双向概率PS-粗糙集的三支决策规则,并分析了三种规则的置信度与错误率的计算方法与性质;其次,定义了决策度量函数与决策损失函数,并基于贝叶斯决策论的最小风险决策规则,给出了参数阈值的计算方法;最后,讨论了此动态模型的决策性质,并通过元素迁入和迁出的实例证明了模型的正确性与可行性.

英文摘要：

In practice,the set is dynamic,including elements moving in and moving out.The bidirectional transfer probabilistic PS-rough sets not only take into account the dynamic characteristics of the set but the rough sets theory of statistical information in the knowledge base.In order to solve the issue of three way decisions under set＇s dynamic changing,the model of three way decisions model based on bidirectional transfer probabilistic PS-rough sets was put forward.Firstly,according to the upper and lower approximation of bidirectional transfer probabilistic PS-rough sets,the probabilistic PS-positive region,negative region and boundary region were got,the rules of three way decisions on the bidirectional transfer probabilistic PS-rough sets were given,and the computing method and its properties of con-fidence coefficient and error rate were analyzed.Then,decision metric function and decision loss function of three-way decision based on bidirectional transfer probabilistic PS-rough sets were defined,and the estimation method of the threshold value based on the minimum risk decision rules of Bayesian decision theory was given.Finally,the decision properties of the dynamic model were discussed;the correctness and feasibility of the model were proved by an example of moving in and moving out of elements.