single-cell-rna-seq

Proposes scKDGM, a framework that uses KAN-guided dynamic graph masked learning and cross-view contrastive learning for clustering single-cell RNA-seq data, achieving state-of-the-art performance on 12 real datasets.