基于核熵成分分析的流式数据自动分群方法

发布时间：2018-02-13 19:05

  本文关键词： 流式细胞术 自动分群 核熵成分分析 K-means算法 余弦相似度　出处：《仪器仪表学报》2017年01期 　论文类型：期刊论文

【摘要】：针对多参数流式细胞数据传统人工分群过程复杂、自动化程度不高等问题,提出了一种基于核熵成分分析(KECA)的自动分群方法。选取对瑞利(Renyi)熵具有最大贡献的特征向量作为投影方向,对数据进行特征提取;设计了一种基于余弦相似度和K-means算法的分类器,并采用一种基于向量夹角的最佳聚类数确定方法,最终获得细胞的分类标签。对实验获得的淋巴细胞免疫表型分析数据进行处理,结果表明,该方法能够实现细胞的快速、自动分群,整体分群准确率能够达到97%以上,操作简单便捷,提高了细胞分析的效率。
[Abstract]:In view of the traditional artificial clustering process of multi-parameter flow cell data, the process is complex and the degree of automation is not high. An automatic clustering method based on Kernel Entropy component Analysis (KECA) is proposed. The feature vector which has the greatest contribution to Rayleigh Renyi entropy is selected as the projection direction to extract the feature of the data. A classifier based on cosine similarity and K-means algorithm is designed. Finally, the classification label of cells was obtained. The experimental data of lymphocyte immunophenotypic analysis were processed. The results show that this method can realize the rapid and automatic grouping of cells, and the accuracy of overall grouping can reach more than 97%. The operation is simple and convenient, and the efficiency of cell analysis is improved.
【作者单位】： 北京信息科技大学光电测试技术北京市重点实验室;
【基金】：教育部"长江学者与创新团队"发展计划(IRT1212) 国家重大科学仪器设备开发专项基金(2011YQ030134) 国家自然科学基金(61605010)项目资助
【分类号】：Q2-3;TP311.13
，

本文编号：1508882