Research map / Federated learning and privacy / Personalized and heterogeneous FL
Federated Learning via Variational Bayesian Inference: Personalization, Sparsity and Clustering
Xu Zhang, Wenpeng Li, Yunfeng Shao, Yinchuan Li, Kaiwen Zhou, Yinchuan Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace point-estimate personalization with Bayesian personalized model (pFedBayes) | personalized FL · CIFAR-10 small, personalized model | accuracy (%): 49.66 -> 61.63 | Improved |
| use pFedBayes instead of strongest baseline on small SVHN | personalized FL · SVHN small, personalized model | accuracy (%): 56.79 -> 62.53 | Improved |
| evaluate pFedBayes at 400 instead of 800 rounds (matching FedAvg communication) (ablation) | pFedBayes · MNIST small, personalized model | accuracy (%): 94.72 -> 94.46 | No clear effect |
| add Bayesian aggregation for the global model on large FMNIST | pFedBayes global model · FMNIST large, global model | accuracy (%): 85.42 -> 83.41 | Hurt |
| add spike-and-slab sparsity (lambda_init 0.5) to pFedBayes (ablation) | pFedBayes · MNIST medium, personalized model | accuracy (%): 97.43 -> 97.60 | Improved |
| add spike-and-slab sparsity (lambda_init 0.5) to pFedBayes (ablation) | pFedBayes · CIFAR-10 medium, personalized model | accuracy (%): 75.21 -> 77.24 | Improved |
| add sparsity (lambda_init 0.5) to pFedBayes (ablation) | pFedBayes · CIFAR-10 medium, global model | accuracy (%): 61.22 -> 55.04 | Hurt |
| increase lambda_init from 0.3 to 0.7 in sFedBayes (ablation) | sFedBayes · CIFAR-10 small, personalized model | accuracy (%): 63.53 -> 69.93 | Improved |
| increase lambda_init from 0.3 to 0.7 in sFedBayes (ablation) | sFedBayes · MNIST medium, personalized model | accuracy (%): 97.54 -> 96.93 | Hurt |
| use multiple cluster priors (cFedBayes) instead of one global prior | pFedBayes · MNIST-MNIST-perp medium, personalized model | accuracy (%): 92.89 -> 94.41 | Improved |
| use multiple cluster priors (cFedBayes) instead of one global prior | pFedBayes · FMNIST-FMNIST-perp medium, personalized model | accuracy (%): 83.77 -> 85.14 | Improved |
| use multiple cluster priors (cFedBayes) instead of one global prior | pFedBayes · CIFAR10-CIFAR10-perp large, global model | accuracy (%): 62.95 -> 68.95 | Improved |
Closest papers by meaning
- A Bayesian Approach for Personalized Federated Learning in Heterogeneous Settings (NeurIPS 2024 · similarity 0.93)
- Harnessing Heterogeneous Statistical Strength for Personalized Federated Learning via Hierarchical Bayesian Inference (ICML 2025 · similarity 0.92)
- Personalized Bayesian Federated Learning with Wasserstein Barycenter Aggregation (NeurIPS 2025 · similarity 0.91)
- Federated Generalised Variational Inference: A Robust Probabilistic Federated Learning Framework (ICML 2025 · similarity 0.89)
- Adaptive Latent-Space Constraints in Personalized Federated Learning (NeurIPS 2025 · similarity 0.89)
- Calibrated One Round Federated Learning with Bayesian Inference in the Predictive Space (AAAI 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
