决策粗糙集风险偏好模型的改进

摘要

决策粗糙集（DTRS）通过引入 Bayes 风险决策理论和三枝决策语义，为不确定知识的获取提供了更可靠的理论依据和语义解释。决策粗糙集的风险偏好模型进一步考虑到决策者的不同风险偏好，使得模型更加贴近实际决策问题。然而，决策粗糙集的风险偏好模型在参数取值范围上有待进一步精确，模型有待进一步完善。提出了一个更准确的决策粗糙集风险偏好模型，并利用 UCI 的信用卡审批数据集进行验证。结果表明，信用评估结果有效且与决策者的风险态度一致。%By bringing in Bayesian decision procedure and three-way decisions,decision-theoretic rough set (DTRS)provided more reliable theoretical basis and semantic interpretation for ac-quisition of uncertain knowledge.The risk preference model based on DTRS gave a further con-sideration on the decision maker's different risk preference,which made the model more suitable for the practical decision making problems.However,the range of values of a parameter in the risk preference model needed to be more accurate.This study proposed a new risk preference model in which the range of values of the parameter was more accurate.This study also took the Credit Approval dataset form UCI to demonstrate the proposed model.The output is not only effective but also in line with the decision makers'risk attitude.