Search papers, labs, and topics across Lattice.
National Yang Ming Chiao Tung University
1
0
3
Biased client selections in federated learning can severely degrade model accuracy, but a new scoring method offers a way to optimize client contributions while preserving privacy.
