Introduction

Predicting exchange-specific latencies requires local trading venue data. Federated learning on edge devices improves models while preserving venue data privacy.

Federated Edge Framework and Training

Models train across trading venues without centralizing data.

Latency Prediction and Applications

Venue-specific latency models improve trading execution and system optimization.

Conclusion

Federated edge learning improves trading venue efficiency while protecting data.