A Hierarchical Incentive Mechanism for Federated Learning
Bibliographic record
Abstract
With the explosive development of mobile computing, federated learning (FL) has been considered as a promising distributed training framework for addressing the shortage of conventional cloud based centralized training. In FL, local model owners (LMOs) individually train their respective local models and then upload the trained local models to the task publisher (TP) for aggregation to obtain the global model. When the data provided by LMOs do not meet the requirements for model training, they can recruit workers to collect data. In this paper, by considering the interactions among the TP, LMOs and workers, we propose a three-layer hierarchical game framework. However, there are two challenges. First, information asymmetry between workers and LMOs may result in that the workers hide their types. Second, incentive mismatch between TP and LMOs may result in a lack of LMOs’ willingness to participate in FL. Therefore, we decompose the hierarchical-based framework into two layers to address these challenges. For the lower-layer, we leverage the contract theory to ensure truthful reporting of the workers’ types, based on which we simplify the feasible conditions of the contract and design the optimal contract. For the upper-layer, the Stackelberg game is adopted to model the interactions between the TP and LMOs, and we derive the Nash equilibrium and Stackelberg equilibrium solutions. Moreover, we develop an iterativeHierarchical-basedUtilityMaximizationAlgorithm (HUMA) to solve the coupling problem between upper-layer and lower-layer games. Extensive numerical experimental results verify the effectiveness of HUMA, and the comparison results illustrate the performance gain of HUMA.
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 | 0.007 | 0.012 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.005 |
| Open science | 0.004 | 0.003 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.006 | 0.001 |
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".