Summary

In this chapter, we discussed the execution of FL systems in detail and how the system will behave according to the interactions between the aggregator and agents. The step-by-step explanation of the FL system behavior based on the outcomes of the console examples guides you to understand the aggregation process of the FedAvg algorithm. Furthermore, the image classification example showed how CNN models are connected to the FL system and how the FL process increases the accuracy through aggregation, although this was not optimized to maximize the training results but simplified to validate the integration using CNN.

With what you have learned in this chapter, you will be able to design your own FL applications integrating the principles and framework introduced in this book, and furthermore, will be able to assess the FL behavior on your own to see whether the whole flow of the FL process and model aggregation is happening correctly and consistently.

In the next...