Sample-weighted fused graph-based semi-supervised learning on multi-view data

Jingjun Bi, Fadi Dornaika

Abstract: Highlights•We deal with deep graph-based semi-supervised classification for multi-view data.•A scheme for consensus graph reconstruction for multi-view and non-graph data.•An additional view based on GCN nodes is integrated into the graph fusion scheme.•A scheme for initializing and updating the weights of labeled samples is proposed.•Experiments are performed on six public multi-view image datasets.