PSPU: Enhanced Positive and Unlabeled Learning by Leveraging Pseudo Supervision
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866913423033368576 |
|---|---|
| author | Wang, Chengjie Xu, Chengming Gan, Zhenye Hu, Jianlong Zhu, Wenbing Ma, Lizhuag |
| author_facet | Wang, Chengjie Xu, Chengming Gan, Zhenye Hu, Jianlong Zhu, Wenbing Ma, Lizhuag |
| contents | Positive and Unlabeled (PU) learning, a binary classification model trained with only positive and unlabeled data, generally suffers from overfitted risk estimation due to inconsistent data distributions. To address this, we introduce a pseudo-supervised PU learning framework (PSPU), in which we train the PU model first, use it to gather confident samples for the pseudo supervision, and then apply these supervision to correct the PU model's weights by leveraging non-PU objectives. We also incorporate an additional consistency loss to mitigate noisy sample effects. Our PSPU outperforms recent PU learning methods significantly on MNIST, CIFAR-10, CIFAR-100 in both balanced and imbalanced settings, and enjoys competitive performance on MVTecAD for industrial anomaly detection. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_06698 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | PSPU: Enhanced Positive and Unlabeled Learning by Leveraging Pseudo Supervision Wang, Chengjie Xu, Chengming Gan, Zhenye Hu, Jianlong Zhu, Wenbing Ma, Lizhuag Computer Vision and Pattern Recognition Machine Learning Positive and Unlabeled (PU) learning, a binary classification model trained with only positive and unlabeled data, generally suffers from overfitted risk estimation due to inconsistent data distributions. To address this, we introduce a pseudo-supervised PU learning framework (PSPU), in which we train the PU model first, use it to gather confident samples for the pseudo supervision, and then apply these supervision to correct the PU model's weights by leveraging non-PU objectives. We also incorporate an additional consistency loss to mitigate noisy sample effects. Our PSPU outperforms recent PU learning methods significantly on MNIST, CIFAR-10, CIFAR-100 in both balanced and imbalanced settings, and enjoys competitive performance on MVTecAD for industrial anomaly detection. |
| title | PSPU: Enhanced Positive and Unlabeled Learning by Leveraging Pseudo Supervision |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2407.06698 |