A hierarchical secure aggregation scheme with arbitrary heterogeneous data assignment achieves optimal two-layer communication loads under information-theoretic security against collusions and dropouts.
Gradient coding: Avoiding stragglers in distributed learning,
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.IT 3years
2026 3representative citing papers
Sparse Gaussian and Expansion-Preserving probabilistic gradient codes achieve BIBD-comparable worst-case robustness while extending feasible system parameters via sparsified random matrices.
A cost-preserving transformation enforces information-theoretic secrecy in distributed computing via null-space augmentation of the allocation matrix and shared randomness injection.
citing papers explorer
-
On the Optimality of Hierarchical Secure Aggregation with Arbitrary Heterogeneous Data Assignment
A hierarchical secure aggregation scheme with arbitrary heterogeneous data assignment achieves optimal two-layer communication loads under information-theoretic security against collusions and dropouts.
-
Probabilistic Gradient Coding via Structure-Preserving Sparsification
Sparse Gaussian and Expansion-Preserving probabilistic gradient codes achieve BIBD-comparable worst-case robustness while extending feasible system parameters via sparsified random matrices.
-
Secure Multi-User Linearly-Separable Distributed Computing
A cost-preserving transformation enforces information-theoretic secrecy in distributed computing via null-space augmentation of the allocation matrix and shared randomness injection.