节点文献
高维数据的2D图单点表示原理
2D graphical representation reflecting high-dimensional observation with sole point
【摘要】 高维数据的单点2D图表示是2D图表示的重要分支。由于其可在单幅多元图中显示多个观察,因此适用于模式识别领域中的特征选择与分类空间形成。本文根据多元图中单点映射的变量数目的不同对单点图表示进行了对比,并分析了其各自的适用范围与优缺点。
【Abstract】 2D graphical representation which reflects high-dimensional observation with sole point is one important branch of 2D graphical representations. The representation is competent for feature selection and class space building as it can display a set of observations in one graph. In this paper, the typical representations are analyzed according to the relationship between the sole point and number of attributes. The advantages, disadvantages of each representation and its applicability are also discussed.
【关键词】 单点图表示;
高维数据;
极坐标映射;
散点图;
星座图;
【Key words】 graph representation with sole point; high-dimensional data; polar mapping; scatter; constellation graph;
【Key words】 graph representation with sole point; high-dimensional data; polar mapping; scatter; constellation graph;
【基金】 国家自然科学基金资助项目(60474065;60504035);河北省自然科学基金资助项目(A1217)
- 【文献出处】 燕山大学学报 ,Journal of Yanshan University , 编辑部邮箱 ,2008年05期
- 【分类号】TP391.41
- 【被引频次】2
- 【下载频次】81