摘要
文章讨论了自组织映射、K平均值聚类和一种有效性测度Silhouette指数。针对基因微阵列的数据特点,考虑到自组织映射的优缺点,设计并实现了一种基于聚类有效性测度的自组织映射和K平均值聚类相结合的一种聚类模型。将该模型运用于公开的结肠基因表达数据集和白血病数据集,实验结果表明该模型是行之有效的。
The paper discusses self-organizing feature map （SOM）,K-means clustering and a cluster validity measure, Silhouette index.Considering the character of microarry and SOM,we design and implement a model associating SOM with K-means clustering based on a cluster validity measure.When we apply the model to the gene expression data of leukaemia and colon,good experimental results are gained.It is proved that the model is effective.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第35期210-212,共3页
Computer Engineering and Applications
