Heterogeneity for the Win: One-Shot Federated Clustering

                    Don Kurian Dennis 1 Tian Li 1 Virginia Smith 1

              Abstract                In the case of federated learning, clustering has found ap-
    In this work, we explore the unique challenges—      plications in client-selection (Cho et al., 2020), personal-
    and opportunities—of unsupervised federated        ization (Ghosh et al., 2020) and exploratory data analysis.
    learning (FL). We develop and analyze a ...