Federated Learning - Opportunities and Application Challenges.

In this paper our intention was to present a brief literature review focused on the latest research of the Federated Learning paradigm in order to identify current research trends, possible future directions of development, and challenges in this area. Federated learning as a new, powerful distributed intelligent paradigm can take on various forms in order to fit a diverse set of problems in a wide range of domains, economy, finance, medicine, agriculture and other industrial sectors. Based on presented research results, several key opportunities for future work can be identified and some emerging are connected to communication costs and performance of federated models trained by different algorithms.