Structured learning for unsupervised feature selection with high-order matrix factorization

作者：

Highlights：

• Propose an efficient convergent algorithm for high-order matrix factorization.

• Construct a unified framework for feature selection and data fusion.

• Present one globally structured learning regularizer via sparse representation.

• Establish a new method for optimization problem with orthogonality constraints.

摘要

•Propose an efficient convergent algorithm for high-order matrix factorization.•Construct a unified framework for feature selection and data fusion.•Present one globally structured learning regularizer via sparse representation.•Establish a new method for optimization problem with orthogonality constraints.

论文关键词：Machine learning,Feature selection,Data fusion,Local learning,Graph Laplacian,High-order matrix factorization

论文评审过程：Received 26 July 2018, Revised 27 July 2019, Accepted 16 August 2019, Available online 29 August 2019, Version of Record 8 September 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.112878