FedProC: A Benchmarking Framework for Federated Time-Series Forecasting

Federated Prognos Chronos – A federated learning framework for time-series forecasting.

📖 Documentation: https://fedproc.readthedocs.io/

---

📚 Awesome Federated Learning & Time Series Forecasting

A curated collection of research papers on federated learning and time series forecasting — searchable, filterable, and updated regularly.

👉 fedproc-lemon.vercel.app — browse 8000+ papers by venue, year, and topic

---

Overview

FedProC is a comprehensive federated learning framework designed specifically for time-series forecasting tasks. It enables distributed machine learning across multiple clients while preserving data privacy.

Star History

Features

• 🔒 Privacy-Preserving: Keeps data local while enabling collaborative learning
• 📈 Time-Series Focused: Optimized for forecasting tasks
• 🚀 Scalable: Supports multiple clients and strategies
• 🧩 Modular: Easy to extend with custom models and strategies

Quick Start

Get started with FedProC in just a few steps:

1. Installation - Set up your environment
2. Usage - Run your first experiment
3. Strategies - Choose your federated learning strategy (54 strategies)

Documentation Structure

• Installation - Installation guide and requirements
• Usage - Basic usage and examples
• Strategies - Available federated learning strategies (54 strategies)
• Datasets - Supported datasets and data preparation (31 datasets)
• Models - Available models and architectures (77 models)
• Augmentations - Custom GPU-native PyTorch augmentations (8 augmentations)
• Losses - Loss functions and metrics (19 losses)
• Optimizers - Optimization methods (9 optimizers)
• Schedulers - Learning rate schedulers (9 schedulers)
• Scalers - Data scaling methods (5 scalers)
• Customization - Extending the framework
• Code Formatting - Development guidelines
• Analysis - Results analysis

Contributing

We welcome contributions! Please see our contributing guidelines for more information.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Citation

If you use this work, please cite:

@article{NGUYEN2026102646,
title = {FedProC: A Benchmarking Framework for Federated Time-Series Forecasting},
journal = {SoftwareX},
volume = {34},
pages = {102646},
year = {2026},
issn = {2352-7110},
doi = {https://doi.org/10.1016/j.softx.2026.102646},
author = {Khoa Nguyen and Taehong Kim},
}