| 织物风格研究中的模糊聚类方法 |
| |
| 引用本文： | 潘宁,严灏景.织物风格研究中的模糊聚类方法[J].东华大学学报(自然科学版),1985(1). |
| |
| 作者姓名： | 潘宁 严灏景 |
| |
| 作者单位： | 华东纺织工学院纺织材料研究室，华东纺织工学院纺织材料研究室 |
| |
| 摘 要： | 本文将模糊聚类方法用于织物风格模式识别,取得了较好的效果。为便于相互比较,本文还将同一问题采用多元统计分析中的几种系统聚类方法进行处理,所得结果与模糊聚类都是比较一致的。本文还利用Marriot的G~2|W|准则,确定了织物风格的最佳分类数目G。这就为进一步建立判别函数,作织物样本风格的判别分析提供了前提和依据。这种样本性能鉴别分类的方法具有客观、合理等特点,可望在许多方面得到应用而不仅仅限于织物风格研究。
|
| 关 键 词： | 模式识别 模糊聚类 分组数目确定 |
FUZZY CLUSTER METHOD USED IN FABRIC HAND ASSESSMENT |
| |
| Pan Ning Yah Haojing.FUZZY CLUSTER METHOD USED IN FABRIC HAND ASSESSMENT[J].Journal of Donghua University,1985(1). |
| |
| Authors: | Pan Ning Yah Haojing |
| |
| Institution: | Pan Ning Yah Haojing |
| |
| Abstract: | 1. Fuzzy cluster analysis is applied to pattern recognition of fabric hand with satisfactory results. 2. For the purpose of comparison, the same problem is also tackled by hierachical cluster methods for multivariables analysis. Both results have been found to coincide. 3. Marriot's G~2|W| criterion can be used to determine the optimum group number of clustering, which, consequently, will make it possible for the discriminating analysis to be performed. 4. The method recommended in the paper is also applicable to other similar practical problems besides fabric hand research. |
| |
| Keywords: | pattern recognition fuzzy cluster group numbers determine |
| 本文献已被 CNKI 等数据库收录！ |
|
