| 处理核磁共振数据的一种改进非负矩阵分解算法 |
| |
| 引用本文： | 乔艳花,张晓琴. 处理核磁共振数据的一种改进非负矩阵分解算法[J]. 太原师范学院学报(自然科学版), 2012, 0(4): 72-74 |
| |
| 作者姓名： | 乔艳花 张晓琴 |
| |
| 作者单位： | 山西大学数学科学学院 |
| |
| 摘 要： | 文章将非负矩阵分解算法（NMF）引入到核磁共振数据处理中，由于NMF对因子矩阵随机初始化会造成解不稳定的问题，论文中用系统聚类法对样本进行聚类，然后取其均值初始化基矩阵，最后进行乘性迭代找出最优解，根据系数矩阵获得最终的分类结果，找出影响其分类的变量．
|
| 关 键 词： | 初始化 系统聚类分析 非负矩阵分解 磁模共振 |
A Modified NMF Algorithm for the Date of NMR |
| |
| Qiao Yanhua Zhang Xiaoqin. A Modified NMF Algorithm for the Date of NMR[J]. Journal of Taiyuan Normal University:Natural Science Edition, 2012, 0(4): 72-74 |
| |
| Authors: | Qiao Yanhua Zhang Xiaoqin |
| |
| Affiliation: | Qiao Yanhua Zhang Xiaoqin(School of Mathematical Sciences,Shanxi University,Taiyuan 030006,China) |
| |
| Abstract: | The NMF algorithm is introduced to the NMR data initialization of factor matrix can cause instability. Firstly,we use sy processing, because random stem clustering method, and then take the mean for the matrix initialization,the multiplicative iterative rules to find the optimal solution. According to the coefficient matrix to obtain the final classification results, and to find out the variables which influent the classification. |
| |
| Keywords: | initialization system clustering method non-negative matrix factorization nuclear magnetic resonance |
| 本文献已被 CNKI 维普 等数据库收录！ |
