Non-linear dimensionality reduction techniques for classification and visualization
Non-linear dimensionality reduction techniques for classification and visualization
复制标题
DOI：
10.1145/775047.775143
复制
发表时间：
2002-07
期刊：
影响因子：
--
通讯作者：
M. Vlachos;C. Domeniconi;D. Gunopulos;G. Kollios;Nick Koudas
中科院分区：
文献类型：
--
作者：
M. Vlachos;C. Domeniconi;D. Gunopulos;G. Kollios;Nick Koudas
在本文中，我们探讨了在二维和三维数据可视化以及分类中使用局部嵌入的问题。我们主张使用它们是因为它们提供了一种从数据的原始维度到较低内在维度的有效映射过程。我们描述了它们如何能够为了可视化目的准确地捕捉用户对高维数据相似性的感知。此外，我们利用这些嵌入所提供的低维映射来开发新的分类技术，并且通过实验表明，分类精度（尽管使用的维度较少）与许多其他分类方法相当。
In this paper we address the issue of using local embeddings for data visualization in two and three dimensions, and for classification. We advocate their use on the basis that they provide an efficient mapping procedure from the original dimension of the data, to a lower intrinsic dimension. We depict how they can accurately capture the user's perception of similarity in high-dimensional data for visualization purposes. Moreover, we exploit the low-dimensional mapping provided by these embeddings, to develop new classification techniques, and we show experimentally that the classification accuracy is comparable (albeit using fewer dimensions) to a number of other classification procedures.