Active learning

Bayesian Batch Active Learning as Sparse Subset Approximation

We reconsider the active learning problem, leveraging advances in Bayesian coresets to relieve the standard myopic assumption.