arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1403.5994stat.ML
收藏

大规模噪声数据鲁棒非负矩阵分解的一阶方法

First Order Methods for Robust Non-negative Matrix Factorization for Large Scale Noisy Data

Jason Gejie Liu, Shuchin Aeron

摘要:

非负矩阵分解(NMF)在可分性假设下是可识别的,在可分性假设下,输入数据矩阵的所有列(或行)都属于仅由这些列(或行)中的少数列(或行)生成的凸锥[1]。然而,在实际应用中,这种可分性假设很难满足。继[4]和[5]之后,在本文中,我们研究了基于线性规划(LP)的重新制定,以定位凸锥的极端光线,但在噪声设置中。此外,为了处理大规模的数据,我们采用一阶方法(FOM)来减轻LP的计算复杂性,这主要是由于大量的约束。我们展示了该算法在真实和合成数据集上的性能。

英文摘要:

Nonnegative matrix factorization (NMF) has been shown to be identifiable under the separability assumption, under which all the columns(or rows) of the input data matrix belong to the convex cone generated by only a few of these columns(or rows) [1]. In real applications, however, such separability assumption is hard to satisfy. Following [4] and [5], in this paper, we look at the Linear Programming (LP) based reformulation to locate the extreme rays of the convex cone but in a noisy setting. Furthermore, in order to deal with the large scale data, we employ First-Order Methods (FOM) to mitigate the computational complexity of LP, which primarily results from a large number of constraints. We show the performance of the algorithm on real and synthetic data sets.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑