Towards Stagewise and Energy-Accuracy-balanced Client Selection and Resource Allocation over Dynamic Federated Learning Networks
Bibliographic record
Abstract
Federated Learning (FL) represents a popular distributed learning architecture that facilitates data privacy by enabling clients (e.g., mobile devices) to train a FL model collaboratively without data sharing. Existing efforts mainly focus on accuracy, delay and energy consumption over a rather stable network with certain clients, with less emphasis on the impact of frequent decision-making processes during model training. Inspired by this, this paper considers a dynamic FL network with uncertain participation of clients, while we jointly optimize client selection and communication resource allocation to achieve a balance between energy cost and FL model accuracy, as proved to be NP-hard. To facilitate timely model training, we propose a prediction-based two-stage asynchronous programming mechanism, which decouples the problem in two subproblems, corresponding to two stages. In particular, the former stage determines some long-term clients which are more stable to join in prior to practical model training process, by estimating the online probability of clients. Then, the latter stage can be implemented by involving some temporary clients as backups when long-term ones are not able to show up. Such a well-designed mechanism offers a unique veiw on FL, while enabling a responsive and cost-effective decision-making process. Comprehensive simulations regarding both IID and non-IID data distributions on MNIST and CIFAR-10 datasets can prove our commendable performance on time efficiency, energy cost and accuracy.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.009 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 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; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".