The authors propose a game-theoretic framework to address privacy concerns in federated learning, considering both defenders and attackers' payoffs. They introduce an oracle to provide lower and upper bounds of payoffs, facilitating the analysis of optimal strategies.
The author proposes AerisAI, a decentralized collaborative AI framework, to address data privacy issues in federated learning by employing homomorphic encryption and differential privacy techniques.
Proposing AerisAI for secure and efficient decentralized collaborative AI with differential privacy.