Small coresets via negative dependence: DPPs, linear statistics, and concentration

Neural Information Processing Systems 

Because sampling is tractable, DPPs are natural candidates for subsampling tasks, such as minibatch selection or coreset construction.