A peer-to-peer learning algorithm that aggregates neighbor models with weights inversely proportional to their loss on the worker's own data is claimed to be Byzantine-resilient with an arbitrary number of adversaries.
Advances in Neural Information Process ing Systems 32 (2019)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Resilient Peer-to-peer Learning based on Adaptive Aggregation
A peer-to-peer learning algorithm that aggregates neighbor models with weights inversely proportional to their loss on the worker's own data is claimed to be Byzantine-resilient with an arbitrary number of adversaries.