m-SNE: Multiview Stochastic Neighbor Embedding
m-SNE: Multiview Stochastic Neighbor Embedding
复制标题
DOI：
10.1007/978-3-642-17537-4_42
复制
发表时间：
2010-11
期刊：
影响因子：
3.7
通讯作者：
Bo Xie;Yang Mu;Dacheng Tao
中科院分区：
文献类型：
--
作者：
Bo Xie;Yang Mu;Dacheng Tao
在许多现实世界的应用中，可以获得不同的特征（或多视图数据），如何在降维中适当地利用它们是一个挑战。简单地将它们连接成一个长向量是不合适的，因为每个视图都有其特定的统计属性和物理解释。在本文中，我们提出了一种多视图随机邻域嵌入（m-SNE），它系统地将异构特征集成为统一的表示，以便基于概率框架进行后续处理。与传统策略相比，我们的方法可以自动学习每个视图的组合系数，以适应其对数据嵌入的贡献。此外，我们学习组合系数的算法以 的速率收敛，这是平滑问题的最佳速率。对合成数据集和真实数据集的实验表明 m-SNE 在数据可视化和图像检索方面的有效性和鲁棒性。
In many real world applications, different features (or multiview data) can be obtained and how to duly utilize them in dimension reduction is a challenge. Simply concatenating them into a long vector is not appropriate because each view has its specific statistical property and physical interpretation. In this paper, we propose a multiview stochastic neighbor embedding (m-SNE) that systematically integrates heterogeneous features into a unified representation for subsequent processing based on a probabilistic framework. Compared with conventional strategies, our approach can automatically learn a combination coefficient for each view adapted to its contribution to the data embedding. Also, our algorithm for learning the combination coefficient converges at a rate of, which is the optimal rate for smooth problems. Experiments on synthetic and real datasets suggest the effectiveness and robustness of m-SNE for data visualization and image retrieval.