Local algorithms for interactive clustering

Pranjal Awasthi, Maria-Florina Balcan, Konstantin Voevodski

Abstract: We study the design of interactive clustering algorithms for data sets satisfying natural stability assumptions. Our algorithms start with any initial clustering and only make local changes in each step; both are desirable features in many applications. We show that in this constrained setting one can still design provably efficient algorithms that produce accurate clusterings. We also show that our algorithms perform well on real-world data.