Joint adaptive manifold and embedding learning for unsupervised feature selection
作者：
Highlights：
• Propose an unsupervised feature selection approach JAMEL with joint adaptive manifold and embedding learning.
• JAMEL can adaptively learn the manifold structure among data according to the distribution density around the cleaned data in the intrinsic space.
• JAMEL embeds the data into a lower-dimensional space to capture the learnt manifold structure, with the aim to help eliminate the redundant and noisy features.
• JAMEL is very efficient for high dimensional data, with its computational complexity linear to the data dimensionality.
• Experimentally show the effectiveness and efficiency of the proposal.
摘要
•Propose an unsupervised feature selection approach JAMEL with joint adaptive manifold and embedding learning.•JAMEL can adaptively learn the manifold structure among data according to the distribution density around the cleaned data in the intrinsic space.•JAMEL embeds the data into a lower-dimensional space to capture the learnt manifold structure, with the aim to help eliminate the redundant and noisy features.•JAMEL is very efficient for high dimensional data, with its computational complexity linear to the data dimensionality.•Experimentally show the effectiveness and efficiency of the proposal.
论文关键词：Unsupervised feature selection,Manifold learning,Embedding learning,Sparse learning
论文评审过程：Received 20 March 2019, Revised 25 September 2020, Accepted 29 October 2020, Available online 4 November 2020, Version of Record 30 January 2021.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107742
