On the Benefits of Public Representations for Private Transfer Learning under Distribution Shift
Public pretraining is a promising approach to improve differentially private model training. However, recent work has noted that many positive research results studying this paradigm only consider in-distribution tasks, and may not apply to settings where there is distribution shift between the pretraining and finetuning data -- a scenario that is likely when finetuning private tasks due to the sensitive nature of the data. In this work, we show empirically across three tasks that even in settings with large distribution shift, where both zero-shot performance from public data and training from scratch with private data give unusably weak results, public features can in fact improve private training accuracy by up to 67\% over private training from scratch. We provide a theoretical explanation for this phenomenon, showing that if the public and private data share a low-dimensional representation, public representations can improve the sample complexity of private training even if it is impossible to learn the private task from the public data alone. Altogether, our results provide evidence that public data can indeed make private training practical in realistic settings of extreme distribution shift.
Code (0)
등록된 구현이 없습니다.
Tasks
regressionTransfer LearningSimilar Papers 제목 키워드 기반
Why Is Public Pretraining Necessary for Private Model Training?
In the privacy-utility tradeoff of a model trained on benchmark language and vision tasks, remarkable improvements have been widely reported with the use of pretraining on publicly available data. This is in part due to …
Transfer LearningPublic policy for management of forest pests within an ownership mosaic
Urban forests provide ecosystem services that are public goods with local (shade) to global (carbon sequestration) benefits and occur on both public and private lands. Thus, incentives for private tree owners to invest i…
ManagementDifferentially Private Prototypes for Imbalanced Transfer Learning
Machine learning (ML) models have been shown to leak private information from their training datasets. Differential Privacy (DP), typically implemented through the differential private stochastic gradient descent algorit…
Transfer LearningPublic-private funding models in open source software development: A case study on scikit-learn
Governments are increasingly funding open source software (OSS) development to support software security, digital sovereignty, and national competitiveness in science and innovation, amongst others. However, little is kn…
TransFR: Transferable Federated Recommendation with Pre-trained Language Models
Federated recommendations (FRs), facilitating multiple local clients to collectively learn a global model without disclosing user private data, have emerged as a prevalent architecture for privacy-preserving recommendati…
Privacy Preserving