KMS Chongqing Institute of Green and Intelligent Technology, CAS
Advancing Non-Negative Latent Factorization of Tensors With Diversified Regularization Schemes | |
Wu, Hao1,2,3; Luo, Xin1,2,4; Zhou, Mengchu5,6,7 | |
2022-05-01 | |
摘要 | Dynamic relationships are frequently encountered in big data and services computing-related applications, like dynamic data of user-side QoS in Web services. They are modeled into a high-dimensional and sparse (HiDS) tensor, which contain rich knowledge regarding temporal patterns. A non-negative latent factorization of tensors (NLFT) model is very effective in extracting such patterns from an HiDS tensor. However, it commonly suffers from overfitting with improper regularization schemes. To address this issue, this article investigates NLFT models with diversified regularization schemes. Six regularized NLFT models, i.e., L-2, L-1, elastic net, log, dropout, and swish-regularized ones, are proposed and carefully investigated. Moreover, owing to their diversified regularization designs, they possess strong model diversity to achieve an effective ensemble. Empirical studies on HiDS QoS tensors from real applications demonstrate that compared with state-of-the-art models, the proposed ones better describe the temporal patterns hidden in an HiDS tensor, thereby achieving significantly higher prediction accuracy for missing data. Moreover, their ensemble further outperforms each of them in terms of prediction accuracy for missing QoS data. |
关键词 | High-dimensional and sparse tensor missing data latent factor analysis temporal pattern non-negativity non-negative latent factorization of tensor regularization ensemble services computing |
DOI | 10.1109/TSC.2020.2988760 |
发表期刊 | IEEE TRANSACTIONS ON SERVICES COMPUTING |
ISSN | 1939-1374 |
卷号 | 15期号:3页码:1334-1344 |
通讯作者 | Luo, Xin(luoxin21@cigit.ac.cn) |
收录类别 | SCI |
WOS记录号 | WOS:000812532400013 |
语种 | 英语 |