Efficient Multi-Task Subset Selection with Distributional Robustness

The core message of this work is to introduce a novel formulation for multi-task submodular optimization that achieves local distributional robustness within the neighborhood of a reference distribution, which assigns importance scores to each task.