FedHPO-B: A Benchmark Suite for Federated Hyperparameter Optimization
This work addresses the problem of comparing HPO methods in federated learning for researchers, but it is incremental as it adapts existing benchmarking concepts to a new domain.
The authors tackled the lack of benchmarks for hyperparameter optimization (HPO) in federated learning (FL) by proposing FedHPO-B, a benchmark suite that includes comprehensive FL tasks and enables efficient evaluations, with results showing it facilitates research and benchmarking of HPO methods in this setting.
Hyperparameter optimization (HPO) is crucial for machine learning algorithms to achieve satisfactory performance, whose progress has been boosted by related benchmarks. Nonetheless, existing efforts in benchmarking all focus on HPO for traditional centralized learning while ignoring federated learning (FL), a promising paradigm for collaboratively learning models from dispersed data. In this paper, we first identify some uniqueness of HPO for FL algorithms from various aspects. Due to this uniqueness, existing HPO benchmarks no longer satisfy the need to compare HPO methods in the FL setting. To facilitate the research of HPO in the FL setting, we propose and implement a benchmark suite FedHPO-B that incorporates comprehensive FL tasks, enables efficient function evaluations, and eases continuing extensions. We also conduct extensive experiments based on FedHPO-B to benchmark a few HPO methods. We open-source FedHPO-B at https://github.com/alibaba/FederatedScope/tree/master/benchmark/FedHPOB.