Fast Discrete Distribution Clustering Using Wasserstein Barycenter with Sparse Support
This addresses a scalability issue in clustering discrete distributions for researchers in fields using weighted bag-of-vectors or histogram descriptors, though it is incremental as it builds on existing D2-clustering methods.
The paper tackles the scalability bottleneck in D2-clustering of discrete distributions by developing a modified Bregman ADMM method to compute approximate Wasserstein barycenters with sparse support, achieving high accuracy at reduced computational cost and demonstrating competitive clustering results across multiple datasets.
In a variety of research areas, the weighted bag of vectors and the histogram are widely used descriptors for complex objects. Both can be expressed as discrete distributions. D2-clustering pursues the minimum total within-cluster variation for a set of discrete distributions subject to the Kantorovich-Wasserstein metric. D2-clustering has a severe scalability issue, the bottleneck being the computation of a centroid distribution, called Wasserstein barycenter, that minimizes its sum of squared distances to the cluster members. In this paper, we develop a modified Bregman ADMM approach for computing the approximate discrete Wasserstein barycenter of large clusters. In the case when the support points of the barycenters are unknown and have low cardinality, our method achieves high accuracy empirically at a much reduced computational cost. The strengths and weaknesses of our method and its alternatives are examined through experiments, and we recommend scenarios for their respective usage. Moreover, we develop both serial and parallelized versions of the algorithm. By experimenting with large-scale data, we demonstrate the computational efficiency of the new methods and investigate their convergence properties and numerical stability. The clustering results obtained on several datasets in different domains are highly competitive in comparison with some widely used methods in the corresponding areas.