1 paper
Patrick Chao, Edgar Dobriban
Contaminations are a key concern in modern statistical learning, as small but systematic perturbations of all datapoints can substantially alter estimation results. Here, we study…