loretanr/dp-gbdt

GBDT learning + differential privacy. Standalone C++ implementation of "DPBoost" (Li et al.). There are further hardened & SGX versions of the code.

29
/ 100
Experimental

This project provides an implementation of differentially private Gradient Boosted Decision Trees (GBDT). It takes structured data as input and produces a GBDT model that can perform regression or binary classification while protecting the privacy of individual data points. This is ideal for data scientists, machine learning engineers, or researchers working with sensitive datasets who need to build predictive models without compromising privacy.

No commits in the last 6 months.

Use this if you need to build machine learning models with strong privacy guarantees, especially when dealing with sensitive data that requires differential privacy.

Not ideal if you need to perform multi-class classification, or if your primary concern is model performance without the specific requirement for differential privacy, as there are still minor privacy issues being resolved.

data-privacy machine-learning-engineering secure-computation predictive-modeling sensitive-data-analysis
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 8 / 25

How are scores calculated?

Stars

9

Forks

1

Language

C++

License

MIT

Last pushed

May 19, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/loretanr/dp-gbdt"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.