{"id":"arxiv/2610.01404","url":"https://arxiv.org/abs/2610.01404","tags":["math.OC"],"score":0,"title":"Constant-Modulus Optimization for Quadratic K-Medoids Graph Clustering","domain":"math.OC","source":"arxiv","authors":["Yutong Zheng","Qingna Li","Wenshun Teng"],"reviews":0,"subject":"math.OC","summary":"Graph clustering aims to partition the vertices of a graph into groups with dense intra-cluster connections and sparse inter-cluster connections. In this paper, we study a quadratic $K$-medoids formulation for graph clustering from the perspective of CM optimization. By identifying the binary feasible set as a discrete CM set, we establish a connection between the quadratic $K$-medoids model and CM optimization. Rather than solving the original binary problem directly, we consider the convex hull of its feasible set and introduce a negative squared-norm penalty to promote extreme-point…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-01","reactions":{}}