General and Robust Communication-Efficient Algorithms for Distributed Clustering
May 3, 2017

As datasets become larger and more distributed, algorithms for distributed clustering have become more and more important. In this work, we present a general framework for designing distributed clustering algorithms that are robust to outliers. Using our framework, we give a distributed approximation algorithm for k-means, k-median, or generally any L_p objective, with z outliers and/or balance constraints, using O(m(k+z)(d+log n)) bits of communication, where m is the number of machines, n is the size of the point set, and d is the dimension. This generalizes and improves over previous work of Bateni et al. and Malkomes et al. As a special case, we achieve the first distributed algorithm for k-median with outliers, answering an open question posed by Malkomes et al. For distributed k-means clustering, we provide the first dimension-dependent communication complexity lower bound for finding the optimal clustering. This improves over the lower bound from Chen et al. which is dimension-agnostic.

Furthermore, we give distributed clustering algorithms which return nearly optimal solutions, provided the data satisfies the approximation stability condition of Balcan et al. or the spectral stability condition of Kumar and Kannan.

Joint work with Nina Balcan and Pranjal Awasthi.