distribution shift – Optimization Online

Learning Optimal Classification Trees Robust to Distribution Shifts

Published: 2023/10/26

(Mixed) Integer Linear Programming, Robust Optimization decision trees, distribution shift, mixed-integer programming, robust machine learning, robust optimization

We consider the problem of learning classification trees that are robust to distribution shifts between training and testing/deployment data. This problem arises frequently in high stakes settings such as public health and social work where data is often collected using self-reported surveys which are highly sensitive to e.g., the framing of the questions, the time … Read more

Robust Generalization despite Distribution Shift via Minimum Discriminating Information

Published: 2021/06/05, Updated: 2021/06/10

Andreas Krause

Daniel Kuhn

Tobias Sutter

Robust Optimization data-driven decision making, distribution shift, distributionally robust optimization, large deviations, principle of minimum discriminating information, stochastic programming

Training models that perform well under distribution shifts is a central challenge in machine learning. In this paper, we introduce a modeling framework where, in addition to training data, we have partial structural knowledge of the shifted test distribution. We employ the principle of minimum discriminating information to embed the available prior knowledge, and use … Read more