Cross-view graph matching for incomplete multi-view clustering

•A novel incomplete multi-view clustering method (CGMIMC) is proposed.•CGMIMC uses the intrinsic structure of each view to complete the missing samples.•CGMIMC constructs the connection graph to preserve the similarity between instances.•CGMIMC adopts the cross-view graph matching to mine the consistent information.•Numerical experiments on several standard datasets are conducted for validation.