Scalability and Efficiency in Federated Learning
Alyan Zaib · 2024
In this chapter, we will study how to make federated learning systems work well when there is an overload of information. We seek to make the systems efficient by using scalability and efficiency in federated learning. We will see that it is like a group of people learning together, sharing their own information. We can set up processes that ensure the model works well, even with vast quantities of data. The research will focus on making these systems work not only with a large amount of data, but also work in a smarter and more efficient way. We will display some techniques on how to improve how these systems interact with each other and use their resources wisely. This will all serve to improve the learning models, like when a group of students work collaboratively. Scalability is particularly complex in conditions where federated learning is positioned across a large number of edge devices, such as smartphones, IoT devices, or edge-computing nodes. The test lies in managing the dissimilarity of these devices, each with changing computational capabilities, energy constraints, and network conditions.