Application of Federated Learning in mobile communication Anti-telecom fraud
Haiou Li, Zhiwen Kang, Guolin Song, Y. Liu, Hong Mei, Xianming Chen, Ming Gu, J. Wang, Geng Chen · 2023
In order to solve the problem that telecom fraud data cannot be centralized for training due to privacy security and data island in the training of telecom fraud model in mobile communication, federated learning can be integrated into the network information interaction to construct an efficient and high-security data model training architecture. The distributed deep learning model training architecture based on federated learning is introduced, and the research status and standard application methods of federated learning are summarized. Based on the model training architecture, an application case test is carried out, and the traditional model training methods are compared. Finally, it is proved that this method is superior to the existing centralized model training method in terms of security and training efficiency under the condition of ensuring network patency.