Fast and direct co-clustering via adaptive anchor graph learning

Jiaqi Nie, Qianyao Qiang

Abstract: Highlights•Propose FDC2 for efficient and direct co-clustering without post-processing.•Learns an adaptive anchor graph to reflect intrinsic data structure accurately.•Eliminates costly eigen-decomposition for better scalability and performance.•Uses a coordinate descent method to solve the discrete indicator matrix.•Achieves superior clustering results with lower computational time.