Tree-Based Federated Learning

Basic Cross-Silo Horizontal Federated Learning PoC using Socket Programming with Decision Tree-based Model Training.

About

Federated Learning (FL) is a groundbreaking approach to collaborative machine learning that prioritizes data privacy by enabling multiple parties to train models without directly sharing raw data. This project focuses on a specific FL paradigm Horizontal Federated Learning.

In this proof of concept, we utilize Walmart's Weekly Sales data to simulate an environment with multiple clients and a server, all hosted locally.Each client represents an independent data silo, with their contributions being aggregated on a central server.

Our implementation employs Socket Programming, a fundamental inter-process communication (IPC) mechanism, to facilitate bidirectional data exchange between distributed clients and the central server.

Run Commands

Running Server

• Open a terminal and navigate to the project directory. Run the following command to start the server: python fl_server.py
• The server will start and begin listening for client connections on 127.0.0.1:12345.

Running Client

• Open multiple terminals (one for each client) and navigate to the project directory in each terminal. Run the following command in each terminal to start a client:python fl_client.py
• Ensure you start the same number of clients as specified in NUM_CLIENTS in fl_server.py. By default, this number is 5.