摘要： 提出一种新的聚类算法，分析待考察属性间的灰关联度，将其转化为属性权值，用于连通分支聚类的距离量度。该算法被用于处理铝电解工业生产中的分类问题。实验结果证明，它在一定程度上克服了欧氏距离的缺陷，能反映属性间的相互影响，提高聚类质量和 性能。

关键词: 灰关联规则, 聚类算法, 欧氏距离

Abstract: This paper proposes a new clustering algorithm which analyzes the gray association value of inspected attribute, transforms them into attribute weight, and applies them in distance calculation of connected components clustering. The algorithm is applied into clustering analysis of categorized problems about cell status in industry production of aluminum electrolysis. Experimental results proves that this algorithm overcomes the defects of the euclidean distance on some extent, reflects the impact of attribute, and improves the quality and performance of clustering.

Key words: gray association rule, clustering algorithm, Euclidean distance

中图分类号: 

• TP391