Skip to content

Latest commit

 

History

History
18 lines (10 loc) · 1.79 KB

README.md

File metadata and controls

18 lines (10 loc) · 1.79 KB

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.