Generalized K-Means Clustering for Apache Spark with Bregman Divergences