Supervised classification in high-dimensional space: Geometrical, statistical, and asymptotical properties of multivariate data
Supervised classification in high-dimensional space: Geometrical, statistical, and asymptotical properties of multivariate data
复制标题
DOI：
10.1109/5326.661089
复制
发表时间：
1998-02-01
期刊：
影响因子：
--
通讯作者：
Landgrebe, DA
中科院分区：
文献类型：
--
作者：
Jimenez, LO;Landgrebe, DA
最近发展了更先进的遥感系统，使得能够在比以前可能的更多的光谱间隔内测量辐射。这种技术的一个例子是AVIRIS系统，它收集220个波段的图像数据。这种高光谱数据的增加的维度极大地增强了数据的信息内容，但对当前用于分析这种数据的技术提出了挑战。人类在三维（3-D）空间中的经验往往会误导我们对高维空间中的几何和统计特性的直觉，这些特性必须指导我们在数据分析过程中的选择。利用欧氏几何和笛卡尔几何，研究了高维空间的性质及其对高维数据及其分析的意义，阐明了高维空间与传统空间的区别。
The recent development of more sophisticated remote-sensing systems enables the measurement of radiation in many more spectral intervals than previously possible. An example of this technology is the AVIRIS system, which collects image data in 220 bands. The increased dimensionality of such hyperspectral data greatly enhances the data information content, but provides a challenge to the current techniques for analyzing such dataHuman experience in three-dimensional (3-D) space tends to mislead our intuition of geometrical and statistical properties in high-dimensional space, properties which must guide our choices in the data analysis process. Using Euclidean and Cartesian geometry, in this paper, high-dimensional space properties are investigated and their implication for high-dimensional data and its analysis is studied to illuminate the differences between conventional spaces and hyperdimensional space.