Technical Note—Data-Driven Chance Constrained Programs over Wasserstein Balls: Exact deterministic reformulation of chance constraintsResearch Paper
A decision can be inexpensive on average yet fail too often to be acceptable. A chance constraint addresses this by bounding the probability of an unsafe outcome. When the distribution of uncertainty is itself unknown, enforcing the constraint only under observed frequencies can overlook outcomes absent from the data. A distributionally robust chance…