PRIVACY PRESERVING MACHINE LEARNING WITH FEDERATED PERSONALIZED LEARNING IN ARTIFICIALLY GENERATED ENVIRONMENT

Abstract

The widespread adoption of Privacy Preserving Machine Learning (PPML) with Federated Personalized Learning (FPL) has been driven by significant advances in intelligent systems research. This progress has raised concerns about data privacy in the artificially generated environment, leading to growing awareness of the need for privacy-preserving solutions. There has been a seismic shift in interest towards Federated Personalized Learning (FPL), which is the leading paradigm for training Machine Learning (ML) models on decentralized data silos while maintaining data privacy. This research article presents a comprehensive analysis of a cutting-edge approach to personalize ML models while preserving privacy, achieved through the innovative framework of Privacy Preserving Machine Learning with Federated Personalized Learning (PPMLFPL). Regarding the increasing concerns about data privacy in virtual environments, this study evaluated the effectiveness of PPMLFPL in addressing the critical balance between personalized model refinement and maintaining the confidentiality of individual user data. According to our results based on various effectiveness metrics, the use of the Adaptive Personalized Cross-Silo Federated Learning with Homomorphic Encryption (APPLE+HE) algorithm for privacy-preserving machine learning tasks in federated personalized learning settings within the artificially generated environment is strongly recommended, obtaining an accuracy of 99.34%.