Ahmed Elkordy’s research interest include privacy preserving, secure and efficient distributed machine learning at the edge. Specifically, Elkordy has been working towards filling the gap between the algorithmic advance of federated learning (FL) with secure model aggregation and its theoretical guarantees. He also has been working in developing efficient algorithms that ensure security against malicious behavior of adversarial users in both the serverless and server-based federated learning settings.
PhD student advised by Salman Avestimehr.