PhD Seminar • Machine Learning | Generative Modeling • Differentially Private Diffusion Models

Please note: This PhD seminar will take place online.

Tim Dockhorn, PhD candidate
David R. Cheriton School of Computer Science

Supervisor: Professor Yaoliang Yu

While modern machine learning models rely on increasingly large training datasets, data is often limited in privacy-sensitive domains. Generative models trained with differential privacy (DP) on sensitive data can sidestep this challenge, providing access to synthetic data instead. However, training DP generative models is highly challenging due to the noise injected into training to enforce DP. We propose to leverage diffusion models (DMs), an emerging class of deep generative models, and introduce Differentially Private Diffusion Models (DPDMs), which enforce privacy using differentially private stochastic gradient descent (DP-SGD).