Legendre decomposition for tensors
Legendre decomposition for tensors
复制标题
DOI：
10.1088/1742-5468/ab3196
复制
发表时间：
2018-02
期刊：
影响因子：
--
通讯作者：
M. Sugiyama;H. Nakahara;K. Tsuda
中科院分区：
文献类型：
--
作者：
M. Sugiyama;H. Nakahara;K. Tsuda
我们提出了一种新的非负张量分解方法，称为勒让德分解，它将输入张量分解为参数的乘法组合。由于发达的信息几何理论，重构张量是唯一的，并且总是最小化从输入张量的KL散度。我们的经验表明，与其他非负张量分解方法相比，勒让德分解可以更精确地重构张量。
We present a novel nonnegative tensor decomposition method, called Legendre decomposition, which factorizes an input tensor into a multiplicative combination of parameters. Thanks to the well-developed theory of information geometry, the reconstructed tensor is unique and always minimizes the KL divergence from an input tensor. We empirically show that Legendre decomposition can more accurately reconstruct tensors than other nonnegative tensor decomposition methods.