Unleashing Pearson Correlation for Faithful Analysis of Biomedical Data
Unleashing Pearson Correlation for Faithful Analysis of Biomedical Data
复制标题
DOI：
10.1007/978-3-642-01805-3_5
复制
发表时间：
2009-01-01
期刊：
影响因子：
--
通讯作者：
Seiffert, Udo
中科院分区：
文献类型：
--
作者：
Strickert, Marc;Schleif, Frank-Michael;Seiffert, Udo
皮尔逊相关系数是生物医学分析中比较的标准之一，具有尚未使用的潜力。通过将皮尔逊相关性转移到自适应相似性度量的框架中并通过利用数学导数的属性来增加实质性价值。这打开了对适用于属性表征、聚类、分类和可视化任务的基于优化的数据模型的访问。现代高通量测量设备对分析大量生物医学数据（包括光谱和高分辨率凝胶电泳图像）产生了很高的需求。在这项研究中，cDNA阵列被认为是感兴趣的数据源。最近的计算方法的表征和分析这些巨大的多维数据集。
Pearson correlation is one of the standards for comparisons in biomedical analyses, possessing yet unused potential. Substantial value is added by transferring Pearson correlation into the framework of adaptive similarity measures and by exploiting properties of the mathematical derivatives. This opens access to optimization-based data models applicable in tasks of attribute characterization, Clustering, classification, and Visualization. Modern high-throughput measuring equipment creates high demand for analysis of extensive biomedical data including spectra and high-resolution gel-electrophoretic images. In this study cDNA arrays are considered as data sources of interest. Recent computational methods are presented for the characterization and analysis of these huge-dimensional data sets.