Information Geometrically Generalized Covariate Shift Adaptation
This work addresses covariate shift adaptation, a key challenge for real-world ML applications where training and test distributions differ, but it appears incremental as it generalizes and improves upon known methods.
The paper tackles the problem of covariate shift in machine learning by unifying existing adaptation methods within an information geometry framework and proposing an efficient parameter search method, achieving better performance than existing methods in numerical experiments.
Many machine learning methods assume that the training and test data follow the same distribution. However, in the real world, this assumption is very often violated. In particular, the phenomenon that the marginal distribution of the data changes is called covariate shift, one of the most important research topics in machine learning. We show that the well-known family of covariate shift adaptation methods is unified in the framework of information geometry. Furthermore, we show that parameter search for geometrically generalized covariate shift adaptation method can be achieved efficiently. Numerical experiments show that our generalization can achieve better performance than the existing methods it encompasses.