中文版 | English
CSpace
Non-Negative Latent Factor Model Based on beta-Divergence for Recommender Systems
Xin, Luo1; Yuan, Ye2,3,4; Zhou, MengChu5,6; Liu, Zhigang2,3; Shang, Mingsheng2,3
2021-08-01
摘要Non-negative latent factor (NLF) models well represent high-dimensional and sparse (HiDS) matrices filled with non-negative data, which are frequently encountered in industrial applications like recommender systems. However, current NLF models mostly adopt Euclidean distance in their objective function, which represents a special case of a beta-divergence function. Hence, it is highly desired to design a beta-divergence-based NLF (beta-NLF) model that uses a beta-divergence function, and investigate its performance in recommender systems as beta varies. To do so, we first model beta-NLF's learning objective with a beta-divergence function. Subsequently, we deduce a general single latent factor-dependent, non-negative and multiplicative update scheme for beta-NLF, and then design an efficient beta-NLF algorithm. The experimental results on HiDS matrices from industrial applications indicate that by carefully choosing the value of beta, beta-NLF outperforms an NLF model with Euclidean distance in terms of accuracy for missing data prediction without increasing computational time. The research outcomes show the necessity of using an optimal beta-divergence function in order to achieve the best performance of an NLF model on HiDS matrices. Hence, the proposed model has both theoretical and application significance.
关键词beta-divergence big data high-dimensional and sparse (HiDS) matrix industrial application learning algorithm non-negative latent factor (NLF) analysis recommender system
DOI10.1109/TSMC.2019.2931468
发表期刊IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
ISSN2168-2216
卷号51期号:8页码:4612-4623
通讯作者Zhou, MengChu(zhou@njit.edu) ; Shang, Mingsheng(msshang@cigit.ac.cn)
收录类别SCI
WOS记录号WOS:000673624500001
语种英语