Parallelized Tensor Train Learning of Polynomial Classifiers

Parallelized Tensor Train Learning of Polynomial Classifiers
复制标题

多项式分类器的并行张量训练学习

DOI:
10.1109/tnnls.2017.2771264
复制
发表时间:
2016-12
影响因子:
10.4
通讯作者:
Ngai Wong
Ngai Wong
中科院分区:
计算机科学1区
文献类型:
--
作者:
Zhongming Chen;Kim Batselier;Johan A K Suykens;Ngai Wong

文献摘要

参考文献

相似文献

在模式分类中,多项式分类器是研究得很好的方法,因为它们能够生成复杂的决策表面。不幸的是,多变量多项式的使用仅限于支持向量机中的内核,因为多项式对于高维问题很快变得不切实际。在本文中,我们有效地克服了灾难的维数采用张量训练(TT)格式表示的多项式分类。基于TTs的结构,提出了两种学习算法,这两种学习算法涉及解决不同的低计算复杂度的优化问题。此外,我们展示了如何将正则化以防止过度拟合和并行化,从而可以使用大型训练集,并将其纳入这些方法中。我们的基于张量的多项式分类器的效率和功效,然后证明了两个流行的数据集美国邮政服务和修改NIST。
In pattern classification, polynomial classifiers are well-studied methods as they are capable of generating complex decision surfaces. Unfortunately, the use of multivariate polynomials is limited to kernels as in support-vector machines, because polynomials quickly become impractical for high-dimensional problems. In this paper, we effectively overcome the curse of dimensionality by employing the tensor train (TT) format to represent a polynomial classifier. Based on the structure of TTs, two learning algorithms are proposed, which involve solving different optimization problems of low computational complexity. Furthermore, we show how both regularization to prevent overfitting and parallelization, which enables the use of large training sets, are incorporated into these methods. The efficiency and efficacy of our tensor-based polynomial classifier are then demonstrated on the two popular data sets U.S. Postal Service and Modified NIST.
DOI: --
发表时间: 2008-09
期刊: --
影响因子: --
作者:
S. Theodoridis;K. Koutroumbas
通讯作者: S. Theodoridis;K. Koutroumbas
DOI: 10.1162/neco_a_00984
发表时间: 2017-07
期刊: Neural Computation
影响因子: 2.9
作者:
J. Suykens
通讯作者: J. Suykens
DOI: --
发表时间: 2009-04
期刊: --
影响因子: --
作者:
R. Brunelli
通讯作者: R. Brunelli
DOI: 10.1007/978-1-4471-0285-4
发表时间: 2012-10
期刊: --
影响因子: --
作者:
Shigeo Abe DrEng
通讯作者: Shigeo Abe DrEng
DOI: 10.1109/21.155943
发表时间: 1992-05-01
期刊: IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS
影响因子: --
作者:
XU, L;KRZYZAK, A;SUEN, CY
通讯作者: SUEN, CY