Tisha-runwal

Tisha-runwal / Personalized-Federated-Learning-for-Privacy--Preserving-and-Scalable-IoT-Driven-Smart-Healthcare

Public

A Personalized Federated Learning (PFL-HCare) framework for IoT healthcare. Features MAML meta-learning, Differential Privacy (RDP), and gradient quantization for efficiency. Includes a React/FastAPI dashboard for real-time monitoring.

doi.org10.1109ICICI65870.2025.11069877 differential-privacy fastapi federated-learning iot-healthcare machine-learning
21
0
100% credibility
Found Apr 09, 2026 at 21 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A dashboard simulating privacy-focused AI training for healthcare IoT devices, visualizing metrics like accuracy, privacy budgets, and communication efficiency across multiple federated learning methods.

How It Works

1
🔍 Discover PFL-HCare

You find this smart healthcare tool on a code-sharing site, promising private AI training for patient data from wearables.

2
🚀 Open the dashboard

Follow easy steps to start the web page on your computer, seeing a sleek interface ready for action.

3
⚙️ Pick your setup

Choose health data like activity tracking or vital signs, number of patient devices, and training style for privacy.

4
▶️ Start learning

Click go, and watch AI models train across pretend devices without sharing any private info.

5
📊 Track live progress

Gaze at colorful charts showing accuracy rising, privacy budget safe, and data savings in real time.

🏆 See winning results

Compare methods to confirm the new approach delivers top accuracy with strong privacy and less network use.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 21 to 21 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is Personalized-Federated-Learning-for-Privacy--Preserving-and-Scalable-IoT-Driven-Smart-Healthcare?

This Python repo delivers a federated learning framework for IoT-driven smart healthcare, training personalized models on edge devices without centralizing sensitive patient data from wearables or monitors. It tackles non-IID medical data distributions, privacy risks, and bandwidth limits using meta-learning adaptation, differential privacy, and gradient compression. Developers get a React/FastAPI dashboard for real-time charts on accuracy convergence, epsilon budgets, and comms overhead, plus CLI scripts to simulate five FL baselines like FedAvg on UCI HAR or synthetic vitals datasets.

Why is it gaining traction?

Unlike bare-bones personalized federated learning github repos, it bundles a production-ready dashboard that visualizes tradeoffs across methods—personalized federated learning a meta-learning approach, with Moreau envelopes, or Gaussian processes—letting you tweak noise multipliers or k-bits live. The Flower integration and Docker setup mean quick local runs without cluster hassle, while 75% bandwidth savings from quantization hook bandwidth-strapped IoT devs. It's a full prototype matching an IEEE paper, with configs for paper reproduction.

Who should use this?

ML engineers prototyping privacy-first health apps on Raspberry Pi fleets, or researchers benchmarking personalized federated learning with theoretical guarantees against Per-FedAvg baselines. Ideal for academics replicating personalized federated conformal prediction with localization on non-IID partitions, or startups building personalized recommendation systems github-style for patient monitoring without GDPR headaches.

Verdict

Grab it for dashboard-driven FL experiments—docs and 49 tests make onboarding fast despite 21 stars and 1.0% credibility score signaling early-stage student work. Scale cautiously beyond sims; add real MIMIC-III for prod.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.