On Noisy Evaluation in Federated Hyperparameter Tuning
This addresses a critical bottleneck for federated learning practitioners by systematically studying and mitigating noise in hyperparameter tuning, though it is incremental as it builds on existing tuning methods.
The paper tackles the problem of noisy evaluation in federated hyperparameter tuning, identifying key sources like client subsampling and heterogeneity, and finds that even small noise reduces state-of-the-art methods to naive baseline performance; it proposes using public proxy data to improve evaluation, achieving a 15% accuracy gain in experiments.
Hyperparameter tuning is critical to the success of federated learning applications. Unfortunately, appropriately selecting hyperparameters is challenging in federated networks. Issues of scale, privacy, and heterogeneity introduce noise in the tuning process and make it difficult to evaluate the performance of various hyperparameters. In this work, we perform the first systematic study on the effect of noisy evaluation in federated hyperparameter tuning. We first identify and rigorously explore key sources of noise, including client subsampling, data and systems heterogeneity, and data privacy. Surprisingly, our results indicate that even small amounts of noise can significantly impact tuning methods-reducing the performance of state-of-the-art approaches to that of naive baselines. To address noisy evaluation in such scenarios, we propose a simple and effective approach that leverages public proxy data to boost the evaluation signal. Our work establishes general challenges, baselines, and best practices for future work in federated hyperparameter tuning.