Add to ORKGAlthough machine learning models trained on massive data have led to break-throughs in several areas, their deployment in privacy-sensitive domains remains limited due to restricted access to data. Generative models trained with privacy constraints on private data can sidestep this challenge, providing indirect access to private data instead. We propose DP-Sinkhorn, a novel optimal transport-based generative method for learning data distributions from private data with differential privacy. DP-Sinkhorn minimizes the Sinkhorn divergence, a computationally efficient approximation to the exact optimal transport distance, between the model and data in a differentially private manner and uses a novel technique for control-ling the bias-variance ...
Generative Adversarial Networks (GANs) are a modern solution aiming to encourage public sharing of d...
A major challenge for machine learning is increasing the availability of data while respecting the p...
Advances in computation have created high demand for large datasets, which in turn has sparked inter...
While modern machine learning models rely on increasingly large training datasets, data is often lim...
Training even moderately-sized generative models with differentially-private stochastic gradient des...
While massive valuable deep models trained on large-scale data have been released to facilitate the ...
The availability of large amounts of informative data is crucial for successful machine learning. Ho...
Leveraging transfer learning has recently been shown to be an effective strategy for training large ...
Differentially private stochastic gradient descent (DP-SGD) is the workhorse algorithm for recent ad...
Training large neural networks with meaningful/usable differential privacy security guarantees is a ...
Machine Learning (ML) is accelerating progress across fields and industries, but relies on accessibl...
In this work, we present an extension to the PyTorch deep learning framework which facilitates diffe...
We study the problem of differentially private (DP) fine-tuning of large pre-trained models -- a rec...
Differentially private data generation techniques have become a promising solution to the data priva...
With the increasing popularity of graph neural networks (GNNs) in several sensitive applications lik...
Generative Adversarial Networks (GANs) are a modern solution aiming to encourage public sharing of d...
A major challenge for machine learning is increasing the availability of data while respecting the p...
Advances in computation have created high demand for large datasets, which in turn has sparked inter...
While modern machine learning models rely on increasingly large training datasets, data is often lim...
Training even moderately-sized generative models with differentially-private stochastic gradient des...
While massive valuable deep models trained on large-scale data have been released to facilitate the ...
The availability of large amounts of informative data is crucial for successful machine learning. Ho...
Leveraging transfer learning has recently been shown to be an effective strategy for training large ...
Differentially private stochastic gradient descent (DP-SGD) is the workhorse algorithm for recent ad...
Training large neural networks with meaningful/usable differential privacy security guarantees is a ...
Machine Learning (ML) is accelerating progress across fields and industries, but relies on accessibl...
In this work, we present an extension to the PyTorch deep learning framework which facilitates diffe...
We study the problem of differentially private (DP) fine-tuning of large pre-trained models -- a rec...
Differentially private data generation techniques have become a promising solution to the data priva...
With the increasing popularity of graph neural networks (GNNs) in several sensitive applications lik...
Generative Adversarial Networks (GANs) are a modern solution aiming to encourage public sharing of d...
A major challenge for machine learning is increasing the availability of data while respecting the p...
Advances in computation have created high demand for large datasets, which in turn has sparked inter...