节点文献
基于偏最小二乘法的信息粒降维及聚类研究
The Research for Information Granule Reduction and Cluster Based on the Partial Least Squares
【摘要】 利用偏最小二乘法,按最大相关性提取出最强解释能力的信息,实现信息粒降维;通过欧氏距离聚类,获取不同层次的信息粒;根据需求选择合适的粒度进行分析.实验结果表明:文中所提方法是可行有效的.
【Abstract】 According to the biggest correlation and the best interpret ability to achieve reducing dimensions based on the partial least squares.To cluster use Euclidean distance,and to obtain different information granular.Analyze the suitable granular in the actual field.It was proved to be feasible and effective after tested with a database.
【关键词】 偏最小二乘法;
信息粒;
降维;
聚类;
【Key words】 partial least squares; information granular; reducing dimensions; cluster;
【Key words】 partial least squares; information granular; reducing dimensions; cluster;
【基金】 国家“973”计划(2010CB530602,2010CB530603);国家“863”计划(2012AA02A609);国家自然科学基金(81160424);江西省自然科学基金(2010GZY0174,20122BAB205083);江西省教育厅课题(GJJ11541)资助项目
- 【文献出处】 江西师范大学学报(自然科学版) ,Journal of Jiangxi Normal University(Natural Science Edition) , 编辑部邮箱 ,2012年05期
- 【分类号】TP311.13
- 【被引频次】3
- 【下载频次】155