Tendi: Tensor Disaggregation from Multiple Coarse Views

Tendi: Tensor Disaggregation from Multiple Coarse Views
复制标题

DOI:
10.1007/978-3-030-47436-2_65
复制
发表时间:
2020-04-17
期刊:
Advances in Knowledge Discovery and Data Mining
影响因子:
--
通讯作者:
Sidiropoulos ND
Sidiropoulos ND
中科院分区:
其他
文献类型:
--
作者:
Almutairi FM;Kanatsoulis CI;Sidiropoulos ND

文献摘要

参考文献

被引文献

相似文献

多维数据出现在各种有趣的应用中,例如,由商店、商品和时间索引的销售数据。通常,观察到的数据是在多个数据单元上聚合的,因此呈现出低分辨率。时间聚合是最常见的,但许多数据集也在其他属性上进行聚合。特别是多维数据有时可在多个粗略视图中获取,这些视图是在不同维度上聚合的——尤其是当数据来自不同机构时。例如,商品销售可以按时间以及根据商店的位置或所属关系对商店分组进行聚合。然而,更细粒度的数据对预测和数据分析有很大益处,这促使人们对数据分解方法越来越感兴趣。在本文中,我们提出了Tendi,这是一个有原则的模型,它能有效地从多个视图中分解在不同维度上聚合的多维(张量)数据。Tendi利用耦合张量分解来融合多个视图,并在实际条件下提供恢复保证。我们还提出了Tendi的一个变体,称为TendiB,它在对聚合机制一无所知的情况下执行分解任务。对来自不同领域的真实数据进行的实验证明了所提出方法的高效性。
Multidimensional data appear in various interesting applications, e.g., sales data indexed by stores, items, and time. Oftentimes, data are observed aggregated over multiple data atoms, thus exhibit low resolution. Temporal aggregation is most common, but many datasets are also aggregated over other attributes. Multidimensional data, in particular, are sometimes available in multiple coarse views, aggregated across different dimensions – especially when sourced by different agencies. For instance, item sales can be aggregated temporally, and over groups of stores based on their location or affiliation. However, data in finer granularity significantly benefit forecasting and data analytics, prompting increasing interest in data disaggregation methods. In this paper, we propose Tendi, a principled model that efficiently disaggregates multidimensional (tensor) data from multiple views, aggregated over different dimensions. Tendi employs coupled tensor factorization to fuse the multiple views and provide recovery guarantees under realistic conditions. We also propose a variant of Tendi, called TendiB, which performs the disaggregation task without any knowledge of the aggregation mechanism. Experiments on real data from different domains demonstrate the high effectiveness of the proposed methods.
DOI: 10.1109/tsp.2018.2876362
发表时间: 2018-12-15
影响因子: 5.4
作者:
Kanatsoulis, Charilaos, I;Fu, Xiao;Ma, Wing-Kin
通讯作者: Ma, Wing-Kin
DOI: 10.1016/s0304-3975(00)00322-4
发表时间: 2001-05-28
影响因子: 1.1
作者:
Kirrinnis, P
通讯作者: Kirrinnis, P
DOI: 10.1109/jstsp.2017.2705581
发表时间: 2017-08-01
影响因子: 7.5
作者:
Almutairi, Faisal M.;Sidiropoulos, Nicholas D.;Karypis, George
通讯作者: Karypis, George
DOI: 10.14778/3236187.3236201
发表时间: 2018-07-01
影响因子: 2.5
作者:
Almutairi, Faisal M.;Yang, Fan;Zadorozhny, Vladimir
通讯作者: Zadorozhny, Vladimir
DOI: 10.1111/jbl.12091
发表时间: 2015-06-01
影响因子: 10.3
作者:
Jin, Yao Henry;Williams, Brent D.;Waller, Matthew A.
通讯作者: Waller, Matthew A.