欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
隶属度转换涉及了众多应用领域,但是,目前已有的隶属度转换算法仍然存在如下问题：给出的转换算法不能揭示指标隶属度中哪部分对目标分类有用,从而导致了指标隶属度中原本对目标分类不起作用的冗余数值也被用于计算目标隶属度.针对已有隶属度转换算法中存在的冗余数据问题,本文从对目标分类的角度设计了一种滤波器,可以识别对目标分类不起作用的冗余的指标隶属度和指标隶属度中的冗余数值,并从指标隶属度中分离出对目标分类起作用的＂有效值＂,参与计算目标隶属度;最终建立实现由指标隶属度到目标隶属度转换的一般算法.应用实例说明了转换的全过程.
英文摘要：
Membership transformation has many practical applications. Existing membership transforming algorithms have some essential problems. For example, they cannot show which parts in the index membership are useful for the objective classification or which parts are of no use. Therefore, the redundant data in the index membership that do not contribute to objective classification are also used to calculate the objective membership. To overcome this problem, we design a kind of filter from the viewpoint of objective classification to identify and remove those redundant index memberships as well as the redundant data in the index membership, and extract the ＂available values＂ for the objective classification. A general transforming algorithm from index membership to the objective membership can be obtained. An application example is introduced to illustrate the transforming process.
同期刊论文项目
同项目期刊论文
位置：