Federated Learning: Privacy-Preserving Machine Learning in Cloud Environments
Abstract
Federated Learning (FL) is a relatively new type of decentralized ML developed to tackle privacy problems inherent to centralized ML methodologies. Thus, FL allows a model to train across multiple distributed devices or edge nodes without exchanging raw data. This approach maintains user privacy since data must not be transmitted to a central location. Only new model parameters are received and recombined to make a global model. This paper looks at the workings of FL to show how it can improve the security of often sensitive data, especially with artificial intelligence solutions based on cloud platforms. The role of federated learning is most prominent in the sectors that work with informational data, including the healthcare and the financial ones. For instance, two hospitals can train models on patient data without compromising their patient's details; similarly, various branches of banks can collectively work out multiple modes of identity theft without compromising the identity of the customers. Therefore, concerning this research, there is clear evidence that FL enhances privacy as well as maintains model efficiency and functionality in different domains. However, there are some problems that can be tied to the use of federated learning. The technical challenges include the communication overhead essential in keeping participants connected, model synchronization that may be a real challenge and encryption that needs to secure the updates made on the model. Also, FL…
Read the paper
Related studies
- Active Learning during Federated Learning for Object Detection
- Fairness-aware federated learning
- Decentralized Federated Learning
- Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner Using Automated Annotation and Federated Learning
- Preservation of Patient Level Privacy: Federated Classification and Calibration Models
- Efficient explainable federated learning
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.