Optimizing Personalized Federated Learning through Adaptive Layer-Wise Learning
This addresses the challenge of optimizing personalized federated learning for real-life deployments with non-IID data, representing an incremental advancement in pFL methods.
The paper tackles the problem of poor accuracy and slow convergence in personalized federated learning (pFL) due to non-IID data by proposing FLAYER, a layer-wise learning method that dynamically adjusts learning rates and selectively uploads parameters, resulting in an average improvement of 5.40% in inference accuracy compared to state-of-the-art methods.
Real-life deployment of federated Learning (FL) often faces non-IID data, which leads to poor accuracy and slow convergence. Personalized FL (pFL) tackles these issues by tailoring local models to individual data sources and using weighted aggregation methods for client-specific learning. However, existing pFL methods often fail to provide each local model with global knowledge on demand while maintaining low computational overhead. Additionally, local models tend to over-personalize their data during the training process, potentially dropping previously acquired global information. We propose FLAYER, a novel layer-wise learning method for pFL that optimizes local model personalization performance. FLAYER considers the different roles and learning abilities of neural network layers of individual local models. It incorporates global information for each local model as needed to initialize the local model cost-effectively. It then dynamically adjusts learning rates for each layer during local training, optimizing the personalized learning process for each local model while preserving global knowledge. Additionally, to enhance global representation in pFL, FLAYER selectively uploads parameters for global aggregation in a layer-wise manner. We evaluate FLAYER on four representative datasets in computer vision and natural language processing domains. Compared to six state-of-the-art pFL methods, FLAYER improves the inference accuracy, on average, by 5.40\% (up to 14.29\%).