Deep multi-view graph clustering network with weighting mechanism and collaborative training

Jing Liu, Fuyuan Cao, Xuechun Jing, Jiye Liang

Abstract: Highlights•The multi-to-one graph autoencoder keeps both diversity and unity of different views.•The view-level weighting mechanisms are designed in both encoder and decoder parts.•The collaborative training for clustering aligns clustering result of each view.•Experiments validate the effect of weighting mechanism and collaborative training.