• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Computational methods in science and engineering: advances in computational science >On the Development of a Hierarchical Fuzzy Classification Framework for Industrial and Environmental Applications
【24h】

On the Development of a Hierarchical Fuzzy Classification Framework for Industrial and Environmental Applications

机译：关于工业和环境应用的层次模糊分类框架的开发

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The aim of this paper is to present the functionality and performance of a general fuzzy classification method that extracts fuzzy rules directly from labeled numerical data using a supervised learning approach. The division of the input space is executed without the use of any a priori knowledge about the data distribution, allows the overlapping of different pattern classes and iteratively aims to reduce the generated classification conflict. The feature space division is affected by two variables that control the extension of the overlapping area and assign an appropriate class to the intersection by affecting the nature and value of the rule weights. The dimensional characteristics of the generated partitions automatically provide the shape and size of the membership functions for each fuzzy feature, and are consequently used for the generation of the equivalent fuzzy rules. This way, the hierarchical input partitioning results to a hierarchically structured rule base that can be used for future classification purposes. Last, an industrial case example for quality control with environmental implications is presented for sake of comparison, when less accuracy is adequate and only offline control is required.
机译：本文的目的是介绍一种通用的模糊分类方法的功能和性能，该方法使用监督学习方法直接从标记的数值数据中提取模糊规则。输入空间的划分是在不使用任何有关数据分布的先验知识的情况下执行的，它允许不同模式类别的重叠，并且迭代地旨在减少所产生的分类冲突。要素空间划分受两个变量影响，这些变量控制重叠区域的扩展并通过影响规则权重的性质和值来为相交指定适当的类。生成的分区的尺寸特征会自动为每个模糊特征提供隶属函数的形状和大小，并因此用于生成等效模糊规则。这样，分层输入分区会生成一个分层结构的规则库，该规则库可用于将来的分类目的。最后，为了进行比较，提出了一个具有环境影响的质量控​​制的工业案例，用于比较精度较低且仅需要离线控制的情况。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号