Generalizable Hyperparameter Optimization for Federated Learning on Non-IID Cancer Images

TL;DR

This study investigates whether hyperparameters optimized on one cancer histopathology dataset can generalize across non-IID federated learning scenarios, proposing a simple heuristic to improve performance.

Contribution

Introduces a dataset-agnostic hyperparameter transfer method using averaging and mode selection, enhancing federated learning on non-IID cancer images.

Findings

The heuristic improves classification accuracy across datasets.

Hyperparameters transferred from one dataset can be effective in others.

Simple averaging of configurations yields competitive results.

Abstract

Deep learning for cancer histopathology training conflicts with privacy constraints in clinical settings. Federated Learning (FL) mitigates this by keeping data local; however, its performance depends on hyperparameter choices under non-independent and identically distributed (non-IID) client datasets. This paper examined whether hyperparameters optimized on one cancer imaging dataset generalized across non-IID federated scenarios. We considered binary histopathology tasks for ovarian and colorectal cancers. We perform centralized Bayesian hyperparameter optimization and transfer dataset-specific optima to the non-IID FL setup. The main contribution of this study is the introduction of a simple cross-dataset aggregation heuristic by combining configurations by averaging the learning rates and considering the modal optimizers and batch sizes. This combined configuration achieves a competitive classification performance.