DORO is a robust outlier refinement of DRO that takes inspiration from its robust statistics. The refined risk function, which prevents DRO from overfitting to potential outliers, intuitively, the new risk function adaptively filters out a small fraction of data with high risk during training, which is potentially caused by outliers.

The post Guide to DORO: Distributional and Outlier Robust Optimization appeared first on Analytics India Magazine.