Domain Generalization Strategy to Train Classifiers Robust to Spatial-Temporal Shift
This work addresses the challenge of making weather prediction models robust to spatial-temporal shifts for improved real-world deployment, though it appears incremental as it builds on existing strategies.
The authors tackled the problem of deep learning-based weather prediction models being vulnerable to spatial-temporal shifts, which hinders real-world application, by proposing a training strategy that combines hyperparameter tuning, data augmentation, and test-time augmentation, achieving first-place performance on the W4C22 Transfer dataset.
Deep learning-based weather prediction models have advanced significantly in recent years. However, data-driven models based on deep learning are difficult to apply to real-world applications because they are vulnerable to spatial-temporal shifts. A weather prediction task is especially susceptible to spatial-temporal shifts when the model is overfitted to locality and seasonality. In this paper, we propose a training strategy to make the weather prediction model robust to spatial-temporal shifts. We first analyze the effect of hyperparameters and augmentations of the existing training strategy on the spatial-temporal shift robustness of the model. Next, we propose an optimal combination of hyperparameters and augmentation based on the analysis results and a test-time augmentation. We performed all experiments on the W4C22 Transfer dataset and achieved the 1st performance.