Abstract Machine learning methods strive to acquire a robust model during the training
process that can effectively generalize to test samples, even in the presence of distribution …

Traditional machine learning paradigms are based on the assumption that both training and
test data follow the same statistical pattern, which is mathematically referred to as …

The current evaluation protocol of long-tailed visual recognition trains the classification
model on the long-tailed source label distribution and evaluates its performance on the …

Distribution shifts—where the training distribution differs from the test distribution—can
substantially degrade the accuracy of machine learning (ML) systems deployed in the wild …

Real-world machine learning deployments are characterized by mismatches between the
source (training) and target (test) distributions that may cause performance drops. In this …

Abstract Machine learning is increasingly used to discover diagnostic and prognostic
biomarkers from high-dimensional molecular data. However, a variety of factors related to …

We challenge a common assumption underlying most supervised deep learning: that a
model makes a prediction depending only on its parameters and the features of a single …

Multi-source domain adaptation (MSDA) methods aim to transfer knowledge from multiple
labeled source domains to an unlabeled target domain. Although current methods achieve …

We introduce the problem of domain adaptation under Open Set Label Shift (OSLS), where
the label distribution can change arbitrarily and a new class may arrive during deployment …

Self-training and contrastive learning have emerged as leading techniques for incorporating
unlabeled data, both under distribution shift (unsupervised domain adaptation) and when it …