Joint graph based embedding and feature weighting for image classification
作者：
Highlights：
• A flexible and discriminant non-linear data embedding is proposed.
• The non-linear model and its regression are simultaneously estimated.
• The proposed regression model explicitly performs feature weighting.
• The model and its kernel variant can be used in supervised and semi-supervised settings.
• Classification performance after embedding is assessed on several public image databases.
摘要
•A flexible and discriminant non-linear data embedding is proposed.•The non-linear model and its regression are simultaneously estimated.•The proposed regression model explicitly performs feature weighting.•The model and its kernel variant can be used in supervised and semi-supervised settings.•Classification performance after embedding is assessed on several public image databases.
论文关键词：Graph-based embedding,Discriminative embedding,Feature weighting,Supervised learning,Semi-supervised learning,Pattern recognition
论文评审过程：Received 6 August 2018, Revised 13 February 2019, Accepted 1 May 2019, Available online 6 May 2019, Version of Record 10 May 2019.
论文官网地址：https://doi.org/10.1016/j.patcog.2019.05.004
