Federated Learning for Societal Impact: Strengths, Open Challenges, and Way Forward

• Olusola Odeyomi (North Carolina A&T State University)

**Abstract**: Federated learning is a machine-learning paradigm that allows multiple clients to collaborate with one another without sharing their data. Federated learning overcomes the limitations of traditional machine learning where clients must upload massive amounts of data to a remote server for centralized learning without privacy considerations. Federated learning has already been adopted for real-world applications, such as Google's Android keyboard and Siri's voice recognition. However, federated learning faces numerous challenges that limit its application across many other domains, such as healthcare and the financial sector. I will be discussing some of these challenges and the current open problems in federated learning. I will recommend possible solutions and their broader impact. **Bio:** Olusola T. Odeyomi is an Assistant Professor of Computer Science at North Carolina A&T State University, North Carolina, United States. He received his Ph.D. degree in Electrical Engineering and Computer Science from Wichita State University, Kansas, United States. He is a member of the IEEE Information Theory Society and the IEEE Computer Science Society. He was a postdoctoral researcher at Howard University and a former lecturer at Obafemi Awolowo University. He won the IEEE International Symposium on Information Theory (ISIT) Video Contest Award in 2021. He was a recipient of the Petroleum Technology Development Fund (PTDF) overseas Ph.D. Scholarship in Nigeria. He is a reviewer for IEEE and ACM journals and conferences. His research interests include machine learning, deep learning, social networks, wireless communication, natural language processing, and bioinformatics. **Note:** See minutes (link at top-right) for references provided by Dr. Odeyomi.