EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Step-Wise Dual Dynamic DPSGD: Enhancing Performance on Imbalanced Medical Datasets with Differential Privacy

Xiaobo Huang, Fang Xie iD

Entropy · 04.04.2026 · 28 (4) · 409

DOI: 10.3390/e28040409 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Entropy, online erschienen am 04.04.2026. Verfasst von Xiaobo Huang und 1 weiteren Autor:innen. Beteiligte Einrichtung: Guangdong Provincial Key Laboratory of IRADS, Beijing Normal-Hong Kong Baptist University, Zhuhai 519087, China. Die Arbeit nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Guangdong Basic and Applied Basic Research Foundation, Guangdong Provincial Key Laboratory IRADS, Guangdong Higher Education Upgrading Plan.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The application of differential privacy in deep learning often leads to significant performance degradation on class-imbalanced medical datasets. Methods such as adding noise to gradients for differential privacy are effective on large datasets, like MNIST and CIFAR-100, but perform poorly on small, imbalanced medical datasets, like HAM10000 and ISIC2019. This is because the imbalanced distribution causes the gradients from the few-shot classes to be clipped, resulting in the loss of crucial information, while the majority classes dominate the learning process. This leads the model to fall into suboptimal solutions early. To address this issue, we propose SDD-DPSGD, which uses a step-wise dynamic exponential scheduling mechanism for noise and clipping thresholds to preserve gradient information. By allocating more privacy budget and employing higher clipping thresholds during the initial training phases, the model can avoid suboptimal solutions and improve its performance. Experiments show that SDD-DPSGD outperforms comparable algorithms on the HAM10000 dataset, and the ISIC2019 dataset.

Schlagwörter

Stichwörter der Autor:innen

convolutional neural networkdeep learningdifferential privacydynamic differentially private mechanismimbalanced dataset

Autor:innen und Institutionen

  • Guangdong Provincial Key Laboratory of IRADS, Beijing Normal-Hong Kong Baptist University, Zhuhai 519087, China.

ORCID-Profile: Fang Xie

Förderung

  • Guangdong Basic and Applied Basic Research Foundation
  • Guangdong Provincial Key Laboratory IRADS
  • Guangdong Higher Education Upgrading Plan

Zitieren

APA

Huang, X., & Xie, F. (2026). Step-Wise Dual Dynamic DPSGD: Enhancing Performance on Imbalanced Medical Datasets with Differential Privacy. Entropy, 28(4), 409. https://doi.org/10.3390/e28040409

Vancouver

Huang X, Xie F. Step-Wise Dual Dynamic DPSGD: Enhancing Performance on Imbalanced Medical Datasets with Differential Privacy. Entropy. 2026;28(4):409. doi:10.3390/e28040409

BibTeX
@article{Huang2026,
  author = {Huang, Xiaobo and Xie, Fang},
  title = {Step-Wise Dual Dynamic DPSGD: Enhancing Performance on Imbalanced Medical Datasets with Differential Privacy},
  journal = {Entropy},
  year = {2026},
  volume = {28},
  number = {4},
  pages = {409},
  doi = {10.3390/e28040409}
}
Kostenlos starten →