Backpropagation Clipping for Deep Learning with Differential Privacy

Stevens, Timothy P.; Ngong, Ivoline C.; Darais, David; Hirsch, Calvin H.; Slater, David; Near, Joseph P.

doi:10.48550/arxiv.2202.05089

Cited by 1 publication

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Tempered sigmoids activations (TSA) control the gradient norm to avoid introducing too much noise. Stevens et al [11] propose the backpropagation clipping (BC) algorithm to replace the gradient clipping operation, which limits the sensitivity of the gradient by clipping the gradient clips each trainable layer's inputs and its upstream gradients. To alleviate clipping bias, Liu et al [12] propose a differentially private learning with grouped gradient clipping (DPL-GGC) method.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Preserving Differential Privacy in Deep Learning Based on Feature Relevance Region Segmentation

Wang

Xie

Tan

et al. 2024

IEEE Trans. Emerg. Topics Comput.

View full text Add to dashboard Cite

In the era of big data, deep learning techniques provide intelligent solutions for various problems in real-life scenarios. However, deep neural networks depend on large-scale datasets including sensitive data, which causes the potential risk of privacy leakage. In addition, various constantly evolving attack methods are also threatening the data security in deep learning models. Protecting data privacy effectively at a lower cost has become an urgent challenge. This paper proposes an Adaptive Feature Relevance Region Segmentation (AFRRS) mechanism to provide differential privacy preservation. The core idea is to divide the input features into different regions with different relevance according to the relevance between input features and the model output. Less noise is intentionally injected into the region with stronger relevance, and more noise is injected into the regions with weaker relevance. Furthermore, we perturb loss functions by injecting noise into the polynomial coefficients of the expansion of the objective function to protect the privacy of data labels. Theoretical analysis and experiments have shown that the proposed AFRRS mechanism can not only provide strong privacy preservation for the deep learning model, but also maintain the good utility of the model under a given moderate privacy budget compared with existing methods.

show abstract

Section: Related Workmentioning

confidence: 99%

“…The cause lies in the fact the more privacy budget means the weaker privacy guarantee, and less noise is injected into the model. [10] 66.20 7.53 BC [11] 74.00 3.64 DPL-GGC [12] 67.11 3.19 ADPSGD [15] 69.63 6.40 DPNASNet [16] 68.33 3.00 AFRRS 76.00 2.00…”

Section: Tablementioning

confidence: 99%