Skip to content

Federated Learning with Differential Privacy + MLflow & Optuna

Notifications You must be signed in to change notification settings

philbeliveau/Retinal_OCT

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Retinal_OCT

Federated Learning with Differential Privacy + MLflow & Optuna

CNN Framework with extensive hyperparamter optimization using optuna and model tracking with mlflow. Best performing model is use as baseline for Federated Learning and Federated Learning with Differential Privacy.

Instructions for use:

  • create your virtual env -> activate
  • pip install -r requirements.txt
    • source folder is the brain
    • data folder has a sample dataset for less intensive experiments
    • helpers: any helper function to prepare the dataset
  • src:
    • 3 folders centralized, federated, and federated_dp holds the codes corresponding to the 3 approaches
    • models - final model parameters
  • run predict.py to evaluate on test data

About

Federated Learning with Differential Privacy + MLflow & Optuna

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%