Family-FL uses family-level aggregation in a three-tier setup with a sub-5KB quantized CNN-LSTM to cut communication by 76.7% versus FedAvg while reaching 91.9% accuracy on MIT-BIH arrhythmia data.
URL https://proceedings.mlsys.org/paper/2021/hash/ 9fe77ac706edb76e4e13183fa9f3b1c8-Abstract.html
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Towards Family-Grouped Hierarchical Federated Learning on Sub-5KB Models: A Feasibility Study of Privacy-Preserving ECG Monitoring for Ultra-Resource-Constrained Wearables
Family-FL uses family-level aggregation in a three-tier setup with a sub-5KB quantized CNN-LSTM to cut communication by 76.7% versus FedAvg while reaching 91.9% accuracy on MIT-BIH arrhythmia data.