A scalable clustering algorithm to approximate graph cuts

Leo Suchan, Housen Li, Axel Munk

Due to their computational complexity, graph cuts for cluster detection and identification are used mostly in the form of convex relaxations. We propose to utilize the original gra…