Incentive mechanism design for Federated Learning with Stackelberg game perspective in the industrial scenario

Federated Learning (FL) is a typical decentralized Machine Learning framework in which clients invest resources to train their local models without sharing their data and then transmit the model parameters to the server for parameter aggregation. Therefore, Incentive Mechanism Design (IMD) is a basic and important research direction in FL that stimulates clients to invest more resources in model training. In this paper, we designed the incentive mechanism as a Stackelberg game in which the server acts as the leader and the clients act as followers. We leveraged modified NSGA-II to find the Nash equilibrium. A real industrial scenario of pre-baked carbon anode quality prediction is applied to verify the high performance of the proposed method.