Evaluation of Hyperparameter-Optimization Approaches in an Industrial Federated Learning System

TL;DR

This paper compares local and global hyperparameter optimization methods in federated learning, focusing on communication efficiency and performance across i.i.d. and non-i.i.d. data distributions.

Contribution

It introduces and evaluates local hyperparameter optimization approaches using grid search and Bayesian methods in federated learning systems.

Findings

Local hyperparameter optimization reduces communication costs.

Bayesian optimization outperforms grid search in accuracy.

Performance varies with data distribution type.

Abstract

Federated Learning (FL) decouples model training from the need for direct access to the data and allows organizations to collaborate with industry partners to reach a satisfying level of performance without sharing vulnerable business information. The performance of a machine learning algorithm is highly sensitive to the choice of its hyperparameters. In an FL setting, hyperparameter optimization poses new challenges. In this work, we investigated the impact of different hyperparameter optimization approaches in an FL system. In an effort to reduce communication costs, a critical bottleneck in FL, we investigated a local hyperparameter optimization approach that -- in contrast to a global hyperparameter optimization approach -- allows every client to have its own hyperparameter configuration. We implemented these approaches based on grid search and Bayesian optimization and evaluated the algorithms on the MNIST data set using an i.i.d. partition and on an Internet of Things (IoT) sensor based industrial data set using a non-i.i.d. partition.