Deep Representation Debiasing via Mutual Information Minimization and Maximization (Student Abstract)

Ruijiang Han, Wei Wang, Yuxi Long, Jiajie Peng

AAAI 2022 pp. 12965-12966

doi:10.1609/AAAI.V36I11.21619 /aaai/2022/han2022aaai-deep/

Abstract

Deep representation learning has succeeded in several fields. However, pre-trained deep representations are usually biased and make downstream models sensitive to different attributes. In this work, we propose a post-processing unsupervised deep representation debiasing algorithm, DeepMinMax, which can obtain unbiased representations directly from pre-trained representations without re-training or fine-tuning the entire model. The experimental results on synthetic and real-world datasets indicate that DeepMinMax outperforms the existing state-of-the-art algorithms on downstream tasks.

PDF AAAI Semantic Scholar

Cite

Text

Han et al. "Deep Representation Debiasing via Mutual Information Minimization and Maximization (Student Abstract)." AAAI Conference on Artificial Intelligence, 2022. doi:10.1609/AAAI.V36I11.21619

Markdown

[Han et al. "Deep Representation Debiasing via Mutual Information Minimization and Maximization (Student Abstract)." AAAI Conference on Artificial Intelligence, 2022.](https://mlanthology.org/aaai/2022/han2022aaai-deep/) doi:10.1609/AAAI.V36I11.21619

BibTeX

@inproceedings{han2022aaai-deep,
  title     = {{Deep Representation Debiasing via Mutual Information Minimization and Maximization (Student Abstract)}},
  author    = {Han, Ruijiang and Wang, Wei and Long, Yuxi and Peng, Jiajie},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2022},
  pages     = {12965-12966},
  doi       = {10.1609/AAAI.V36I11.21619},
  url       = {https://mlanthology.org/aaai/2022/han2022aaai-deep/}
}