Skip to content

Extension to experiments made in the original PrivacyFL code for MNIST dataset with skin segmentation dataset from UCI Machine Learning Repository.

License

Notifications You must be signed in to change notification settings

thisisabhash/PrivacyFL

 
 

Repository files navigation

Federated Learning: What, Why & How?

This is a fork of the repository at https://github.com/vaikkunth/PrivacyFL and contains changes made for the project in my course CS839 - Data Management for Machine Learning at UW-Madison. I am truly grateful to the authors to have provided us a cool framework to experiment with Federated Learning.

This was a project for CS839 - Data Management for Machine Learning at UW-Madison. We extended the experiments made in the original code for MNIST dataset to a skin segmentation dataset from UCI Machine Learning Repository at http://archive.ics.uci.edu/ml/datasets/Skin+Segmentation.

The steps to run the code are the same as in the original repository. The skin dataset is already downloaded and included in the src folder. In case, a file not found error is observed, please make sure that the file Skin_NonSkin.txt is in the same folder as the source code.

Project Report - https://github.com/thisisabhash/PrivacyFL/blob/master/CS839_Federated_Learning_Final_Report.pdf

Citation: Rajen Bhatt, Abhinav Dhall, 'Skin Segmentation Dataset', UCI Machine Learning Repository

Abstract

Federated learning is emerging as a new technique enabling distributed clients to collaborate over a shared Machine Learning model while keeping their data localized, by sharing trained weights instead. Implementing and analyzing a federated learning setup still remains a challenge, with mul- tiple parameters and configurations. In this paper, we aim to do a two-part study. First, we aim to do a bird’s eye view of the federated learning land- scape, and its main challenges. Second, we aim to extend a federated learning simulation system, towards a highly scalable cluster-based environ- ment, to help get a better sense of its advantages in particular machine learning applications.

About

Extension to experiments made in the original PrivacyFL code for MNIST dataset with skin segmentation dataset from UCI Machine Learning Repository.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%