Search papers, labs, and topics across Lattice.
Affiliation:
2
0
3
Adding noise to shared anchor representations instead of private data significantly enhances learning accuracy while maintaining privacy in collaborative settings.
Kernel methods can unlock more accurate and privacy-preserving data collaboration by non-linearly integrating obfuscated representations of decentralized datasets.
