Speaker: Vahideh Hayyolalam – Koç University, Department of Computer Engineering

In this seminar, the real-world challenges of federated learning—including communication bottlenecks, client heterogeneity, and slow convergence—will be addressed. The talk will explore how more efficient, reliable, and scalable distributed artificial intelligence systems can be developed using metaheuristic algorithms, particularly Black Widow Optimization–based methods such as FedBWO and MetaFed.

The presentation will also cover recent developments in optimization-oriented federated learning approaches and highlight future research opportunities in this field.