Fed-Reputed: Reputation-Aware Client Selection in Hierarchical Federated Learning for Consumer Electronics
Bibliographic record
Abstract
Federated Learning (FL) enables collaborative model training across distributed consumer devices without requiring raw data to be centralized—an essential feature for privacy-critical applications such as biomedical monitoring and mobile health diagnostics. However, real-world deployments often face challenges in client selection, especially in the presence of misbehaved clients. Straggler clients with poor connectivity or low resources delay convergence, while malicious clients pose threats to model integrity by injecting poisoned updates. These problems are amplified in dynamic, heterogeneous environments such as those involving consumer biomedical devices. To address this, we propose <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Fed-Reputed</i>, a reputation-aware client selection framework tailored for Hierarchical Federated Learning (HFL). Unlike existing reputation-based approaches that suffer from herding effects, cold-start limitations, and imbalanced classification datasets, Fed-Reputed integrates a modified Bellman equation within a Deep Q-Learning framework. This formulation guides client selection using an Imbalanced Classification Markov Decision Process (ICMDP), leveraging both device capability and historical behavior. Extensive simulations using MNIST and FMNIST datasets under varying percentages of straggler and malicious clients demonstrate that Fed-Reputed achieves up to 50% higher global model accuracy and 1.7 times faster convergence compared to state-of-the-art selection methods. Moreover, it significantly improves the detection of misbehaving clients without sacrificing fairness or scalability. These results highlight the potential of Fed-Reputed to advance secure, scalable, and personalized Federated Learning in health-sensitive consumer environments.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.006 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.008 | 0.001 |
| Research integrity | 0.001 | 0.006 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".