Privacy Preserving machine Learning with Federated personalises Learning in Artificially Generated Environment

With the great progress of research on intelligent systems, the application of Privacy Preserving Machine Learning (PPML) with Federated Personalised Learning (FPL) has been popular. This advancement has resulted in a greater awareness of the importance of privacy-protecting technologies as more people are concerned about data privacy in the artificially generated environment. Federated Personalised Learning (FPL) is gaining tremendous popularity rapidly. FPL is the best way to train machine learning (ML) models on decentralised data silos, while protecting data privacy. In this work we present a comprehensive study of Privacy Preserving Machine Learning with Federated Personalised Learning (PPMLFPL), a novel framework for personalising machine learning models while protecting privacy. In this paper, we explore the potential of PPMLFPL to protect user data privacy and enable personalised model improvement, a topic of increasing importance in virtual settings. The effectiveness metrics used in our work show that the Adaptive Personalised Cross-Silo Federated Learning with Homomorphic Encryption (APPLE+HE) algorithm is a good candidate for privacy-preserving machine learning tasks in the context of federated personalised learning in a synthetic environment, with an accuracy of 99.34%.