FMS (Federated Model as A Service) For Healthcare: An Automated Secure-Framework for Personalized Recommendation System
* Corresponding author
The Healthcare sector has been emerging on the platform of data science. And data scientists are often using machine learning techniques based on historical data to create models, make predictions or recommendations. This paper aims to provide background and information for the community on the benefits and variants of Federated Learning (F.L.) with other technologies for medical applications and highlight key considerations and challenges of F.L. implementation in the digital health background. With this FMaaS, we envisage a future for digital federated health. We hope to empower and raise awareness about the environment and fog computing to provide a more secure and better-analyzing environment. The AutoML framework is used to generate and optimize machine learning models using automatic engineering tools, model selection, and hyperparameter optimization on fog nodes. Thus, making the system more reliable and secure for each individual by preserving privacy at their end devices. And this will lead to a personalized recommendation system for each individual associated with this framework by deploying the Model to their devices for on-device inferences through the concept of differential private Model averaging. With this framework, users don't have to compromise with privacy, and all their sensitive data will be secure on their end devices.
Akshay Saini, Krishnan Ramanathan. FMS (Federated Model as A Service) For Healthcare: An Automated Secure-Framework for Personalized Recommendation System. Cardiometry; Issue 20; November 2021; p.70-78; DOI: 10.18137/cardiometry.2021.20.7078; Available from: https://www.cardiometry.net/issues/no20-november-2021/federated-model-service