A Machine Learning Model to Protect Privacy Using Federal Learning with Homomorphy Encryption
Rejuwan Shamim, Md Arshad, Dr. Vinay Pandey · International Journal for Research in Applied Science and Engineering Technology · 2022
Abstract: Machine learning technologies have a marvelous success in emancipating real-world Artificial Intelligence applications. But still, vast numbers of sensitive data are produced every second time in different forms. This data can be in the form of health records, shopping records, internet searching records, mobile and laptop activities, and so on. This data can be used to train our Machine learning /Deep learning models to make Artificial intelligence-based technologies better than their previous generation. However, in today’s world, one of the significant challenges that need to be a concern in machine learning is regarding data breaches while training the model. Since federated learning trains machine learning algorithms in various devices or servers without sharing sample data. This paper discusses the framework of federated learning and homomorphic encryption and how both frameworks work together so that the outcoming data will be more precious and accurate without bothering data breaches. Later, we focus on its futuristic applications in various fields to improve technology.