Learning matrix factorization with scalable distance metric and regularizer

Shiping Wang, Yunhe Zhang, Xincan Lin, Lichao Su, Guobao Xiao, William Zhu, Yiqing Shi

Abstract: Highlights•Propose a learnable method to approximate several popular matrix factorizations.•Construct an easy-built end-to-end architecture for low-rank learning.•Provide a new perspective for computing hierarchical low-rank matrix factors.•Validate its superiority to matrix factorizations and deep clustering approaches.