Saved in:
| Main Author: | |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2503.05803 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913725370335232 |
|---|---|
| author | Gupta, Yash |
| author_facet | Gupta, Yash |
| contents | Federated Learning often relies on sharing full or partial model weights, which can burden network bandwidth and raise privacy risks. We present a loss-based alternative using distributed mutual learning. Instead of transmitting weights, clients periodically share their loss predictions on a public test set. Each client then refines its model by combining its local loss with the average Kullback-Leibler divergence over losses from other clients. This collaborative approach both reduces transmission overhead and preserves data privacy. Experiments on a face mask detection task demonstrate that our method outperforms weight-sharing baselines, achieving higher accuracy on unseen data while providing stronger generalization and privacy benefits. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_05803 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Federated Learning Framework via Distributed Mutual Learning Gupta, Yash Machine Learning Artificial Intelligence Federated Learning often relies on sharing full or partial model weights, which can burden network bandwidth and raise privacy risks. We present a loss-based alternative using distributed mutual learning. Instead of transmitting weights, clients periodically share their loss predictions on a public test set. Each client then refines its model by combining its local loss with the average Kullback-Leibler divergence over losses from other clients. This collaborative approach both reduces transmission overhead and preserves data privacy. Experiments on a face mask detection task demonstrate that our method outperforms weight-sharing baselines, achieving higher accuracy on unseen data while providing stronger generalization and privacy benefits. |
| title | Federated Learning Framework via Distributed Mutual Learning |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2503.05803 |