Open Federated Learning Extended (OpenFL-extended) - An Open-Source extension of OpenFL supporting Distributed Bagging and Boosting of any ML model

Developed and maintained by Gianluca Mittone (gianluca.mittone@unito.it), Department of Computer Science, University of Turin. Member of the ALPHA group (Parallel and Distributed Computing).

This software offers the possibility to experiment with basic Distributed Bagging and the Distributed Boosting method called AdaBoost.F developed by Polato et al.

[Polato, Mirko, Roberto Esposito, and Marco Aldinucci. "Boosting the federation: Cross-silo federated learning without gradient descent." 2022 International Joint Conference on Neural Networks (IJCNN). IEEE, 2022.]

Installation

You can simply install OpenFL-extended by cloning the distributed_boosting branch of this repo and running pip install . inside the downloaded folder:

$ git clone -b distributed_boosting https://github.com/Giemp95/openfl-extended.git
$ cd openfl-extended
$ pip install .

Getting Started

The quickest way to test OpenFL-extended is to execute the tutorials available in the boosting-examples folder.