Targeted active learning for probabilistic models

Re-assign community