Computational modeling and machine learning (ML) have led to an acceleration of scientific innovation in diverse areas, ranging from drug design to robotics to material science.

Bayesian black-box function paradigmknowledg expected [33,25].

"one-shot" formulation of the Knowledge Gradient, enabled by a combination of

Monte Carlo (MC) integration is the de facto method for approximating the predictive distribution of Bayesian neural networks (BNNs).

Supplementary Materials for "RecursiveMix: Mixed Learning with History" In this section, we review works that "Learning with History" in detail. We then demonstrate a detailed review of existing representative approaches according to the above-mentioned elements. We adopt the popular PVT architecture in our experiments. The COCO datasets contain annotations in 80 categories, with over 1.5 million object instances. The ADE20K dataset contains 25K images annotated with 150 object categories.

Machine learning models often perform poorly under subpopulation shifts in the data distribution. Developing methods that allow machine learning models to better generalize to such shifts is crucial for safe deployment in real-world settings. In this paper, we develop a family of group-aware prior (GAP) distributions over neural network parameters that explicitly favor models that generalize well under subpopulation shifts. We design a simple group-aware prior that only requires access to a small set of data with group information and demonstrate that training with this prior yields state-of-the-art performance -- even when only retraining the final layer of a previously trained non-robust model. Group aware-priors are conceptually simple, complementary to existing approaches, such as attribute pseudo labeling and data reweighting, and open up promising new avenues for harnessing Bayesian inference to enable robustness to subpopulation shifts.