【摘要】 语言特征存在于许多实际应用中,如模糊数知识系统、语义检索、模糊人脸识别等.为解决这类具有语言特征(模糊数)对象的识别/分类问题,提出了基于广义Choquet模糊积分的分类方法与模型.该方法不仅可以解决数值特征对象识别/分类问题,同时为具有模糊数特征的对象识别/分类问题提供了手段,从而扩大了原有分类模型的适用领域.模糊人脸识别模拟实验证明,在语言特征环境下,该分类模型有效地实现了语言模糊对象的近似辨识.更多还原

【Abstract】 Linguistic features exist in many practical fields, such as fuzzy number knowledge system, semantic retrieval, fuzzy face recognition, and so on. This paper presents a new classification method and model based on generalized Choquet fuzzy integral for classifying objects with linguistic features (or fuzzy numbers). It can recognize and classify objects with not only numerical features, but also fuzzy numbers, which extends application fields of existed methods. The simulated experiment on fuzzy face recognition show that it efficiently implements approximate identification of objects whose feature values are linguistic values.更多还原