conftrace_

2026 JMLR JMLR 2026

Abstract Gradient Training: A Unified Certification Framework for Data Poisoning, Unlearning, and Differential Privacy

Abstract

Philip Sosnin, Matthew Wicker, Josh Collyer, Calvin Tsay; 27(128):1−58, 2026.

🌉 Interdisciplinary Bridge - Artificial Intelligence and Security & Privacy

Authors

Philip Sosnin, Matthew Wicker, Josh Collyer, Calvin Tsay

Topics

Security & Privacy > Differential Privacy Security & Privacy > Privacy Machine Learning > Learning Types > Machine Unlearning

Keywords

differential privacy data poisoning machine unlearning certification framework gradient training

More from JMLR 2026

All JMLR 2026 papers →

Approximation-Free Differentiable Oblique Decision Trees

Underdamped Langevin MCMC with third order convergence

Mixing times of data-augmentation Gibbs samplers for high-dimensional probit regression

Doubly Debiased Robust Subsampling for Transfer Learning

Learning to Play Two-Player Perfect-Information Games without Knowledge

Graph-based Clustering Revisited: A Relaxation of Kernel k-Means Perspective