What is ensemble learning with federated learning?

Federated ensemble learning combines models trained locally on decentralized data without sharing raw data.

In federated learning, data remains on local devices due to privacy constraints. Instead of sharing data, each client trains a local model. Ensemble methods aggregate these local models by averaging weights, predictions, or using a meta-model. This preserves privacy while still benefiting from distributed learning. Challenges include non-IID data, communication cost, and model heterogeneity.