.png){kind=icon}
.png){kind=icon}
Return
Deep data representation with feature propagation for semi-supervised learning
.png){kind=icon}
PRE
AI
DOI：10.1007/s13042-022-01701-9.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Graph-based embedding has attracted much attention in the fields of machine learning and pattern recognition. It is becoming an indispensable tool for data representation. It can be useful for all types of learning: unsupervised, semi-supervised, and supervised. In this correspondence, we present a graph-based, deep and flexible method for data representation with feature propagation. The presented framework ensures several desired features such as graph-based regularization, a flexible embedding model, graph-based feature propagation, and a deep learning architecture. The model can be learned layer by layer. In each layer, the nonlinear data representation and the unknown convolved data based regression are jointly estimated with a closed-form solution. We evaluate the proposed system on semi-supervised classification tasks using six public image datasets. These experiments demonstrate the effectiveness of the presented framework, which compares favorably to many competing semi-supervised approaches.
Keywords:
Graph-based projection
Flexible embedding
Deep architecture
Neighborhood feature propagation
Semi-supervised learning
Journal
IF:
2.7
Papers:
3.1K
Citations:
5.6K