Predict then Interpolate: A Simple Algorithm to Learn Stable Classifiers
This addresses the challenge of robust machine learning for applications where data distributions shift, offering a novel method for improving generalization.
The paper tackles the problem of learning stable classifiers across environments by proposing the Predict then Interpolate algorithm, which outperforms IRM by 23.85% on synthetic environments and 12.41% on natural environments in text and image classification tasks.
We propose Predict then Interpolate (PI), a simple algorithm for learning correlations that are stable across environments. The algorithm follows from the intuition that when using a classifier trained on one environment to make predictions on examples from another environment, its mistakes are informative as to which correlations are unstable. In this work, we prove that by interpolating the distributions of the correct predictions and the wrong predictions, we can uncover an oracle distribution where the unstable correlation vanishes. Since the oracle interpolation coefficients are not accessible, we use group distributionally robust optimization to minimize the worst-case risk across all such interpolations. We evaluate our method on both text classification and image classification. Empirical results demonstrate that our algorithm is able to learn robust classifiers (outperforms IRM by 23.85% on synthetic environments and 12.41% on natural environments). Our code and data are available at https://github.com/YujiaBao/Predict-then-Interpolate.