The codes in this warehouse are used for learning purposes. We hope to apply the deep learning approach for the enhancement of tabular data imbalance, and use the contrastive learning method to improve the enhancement effect.
Our model code has learned from TabDiff and MoCo v3. If you want to understand the more fundamental principles and details, you can visit the following link:
TabDiff：https://proceedings.iclr.cc/paper_files/paper/2025/hash/5c882988ce5fac487974ee4f415b96a9-Abstract-Conference.html
MoCo v3：https://ieeexplore.ieee.org/abstract/document/9711302/