A Comprehensive Survey of Federated Learning: Privacy, Personalization, and Future Horizons
Abstract
Federated Learning (FL) is a decentralized machine learning method that enables collaborative model training on various devices or entities without the need to share raw data, thus addressing privacy, security, and regulatory hurdles imposed by industries care, finance and Internet of Things, etc. FL was initially conceived as a system for distributed training and later generalized by the FedAvg algorithm, it has become one of the core technologies of privacy, preserving AI. Research in this field can be roughly divided into five themes: optimization techniques, privacy, preserving strategies, data heterogeneity management, personalization approaches and adversarial attacks resistance. The major achievements are the algorithms for improved convergence in heterogeneous environments as well as the techniques for overcoming the problems caused by non, independent and non, identically distributed data. The publications now focus on real, world issues, methodological trade, offs, and problems such as communication overheads and attack robustness that remain unsolved. Popular research themes explore the integration of FL with blockchain, edge computing and quantum technologies to enhance security, scalability and efficiency.
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.