Maximal Linear Embedding for Dimensionality Reduction

Maximal Linear Embedding for Dimensionality Reduction
复制标题

用于降维的最大线性嵌入

DOI:
10.1109/tpami.2011.39
复制
发表时间:
2011-09
期刊:
IEEE Trans. on Pattern Analysis and Machine Intelligence
影响因子:
--
通讯作者:
Shufu Xie, Shiguang Shan, Xilin Chen, Jie Chen
Shufu Xie, Shiguang Shan, Xilin Chen, Jie Chen
中科院分区:
其他
文献类型:
--
作者:
Shufu Xie, Shiguang Shan, Xilin Chen, Jie Chen

文献摘要

参考文献

被引文献

相似文献

在过去的几十年里,降维在计算机视觉和模式分析中得到了广泛的应用。提出了一种简单有效的非线性降维算法--最大线性嵌入算法。MLE学习参数映射来恢复单个全局低维坐标空间,并为流形产生等距嵌入。受几何直观的启发,我们引入了一个合理的定义,局部线性补丁,最大线性补丁(MLP),它寻求最大化的线性保持的局部邻域。首先将输入数据分解为局部线性模型的集合,每个模型描述一个MLP。然后将这些局部模型对齐到全局坐标空间中,这是通过将MDS应用于一些随机选择的地标来实现的。所提出的对齐方法,称为基于地标的全局对齐(LGA),可以有效地产生一个封闭的形式的解决方案,没有局部最优的风险。它只涉及一些小规模的特征值问题,而大多数以前的对齐技术采用耗时的迭代优化。与传统的方法,如ISOMAP和LLE相比,我们的MLE产生一个显式的模型的观测数据的内在变化模式。在人工数据和真实的数据上的实验结果表明了该算法的有效性。
Over the past few decades, dimensionality reduction has been widely exploited in computer vision and pattern analysis. This paper proposes a simple but effective nonlinear dimensionality reduction algorithm, named Maximal Linear Embedding (MLE). MLE learns a parametric mapping to recover a single global low-dimensional coordinate space and yields an isometric embedding for the manifold. Inspired by geometric intuition, we introduce a reasonable definition of locally linear patch, Maximal Linear Patch (MLP), which seeks to maximize the local neighborhood in which linearity holds. The input data are first decomposed into a collection of local linear models, each depicting an MLP. These local models are then aligned into a global coordinate space, which is achieved by applying MDS to some randomly selected landmarks. The proposed alignment method, called Landmarks-based Global Alignment (LGA), can efficiently produce a closed-form solution with no risk of local optima. It just involves some small-scale eigenvalue problems, while most previous aligning techniques employ time-consuming iterative optimization. Compared with traditional methods such as ISOMAP and LLE, our MLE yields an explicit modeling of the intrinsic variation modes of the observation data. Extensive experiments on both synthetic and real data indicate the effectivity and efficiency of the proposed algorithm.
DOI: 10.1201/b10345-5
发表时间: 2010-11
期刊: Encyclopedia of Autism Spectrum Disorders
影响因子: --
作者:
Kim-Anh Lê Cao;Z. Welham
通讯作者: Kim-Anh Lê Cao;Z. Welham
DOI: 10.1201/9781315220413-4
发表时间: 2018-10
期刊: Handbook of Neural Network Signal Processing
影响因子: --
作者:
Klaus-Robert Müller;S. Mika;Koji Tsuda;Koji Schölkopf
通讯作者: Klaus-Robert Müller;S. Mika;Koji Tsuda;Koji Schölkopf
DOI: --
发表时间: 2002
影响因子: 0.7
作者:
V. Silva;J. Tenenbaum
通讯作者: V. Silva;J. Tenenbaum
DOI: 10.1109/tpami.2005.55
发表时间: 2005-03-01
影响因子: 23.6
作者:
He, XF;Yan, SC;Zhang, HJ
通讯作者: Zhang, HJ
DOI: 10.1109/tpami.2006.166
发表时间: 2006-08
影响因子: 23.6
作者:
J. Verbeek
通讯作者: J. Verbeek