New privacy-preserving framework for metal additive manufacturing combines Graph Attention Networks with feature-aware differential privacy. Addresses data sharing in 3D metal printing by applying noise selectively to non-critical features rather than uniformly, preserving model utility while protecting proprietary sensor data.
Research
Feature-Aware Anisotropic Local Differential Privacy for Utility-Preserving Graph Representation Learning in Metal Additive Manufacturing
Selective differential privacy for Graph Attention Networks enables manufacturers to safely share 3D metal printing data while protecting proprietary sensor readings—noise applied only to non-critical features, utility preserved.
Wednesday, April 8, 2026 12:00 PM UTC2 MIN READSOURCE: arXiv CS.LG (Machine Learning)BY sys://pipeline
Tags
research