Optimising privacy protection for cross-border data flow based on federated learning

As cross-border data flows continue to expand, protecting people's privacy has become a major barrier to international cooperation.This research provides a privacy-preserving optimisation technique based on federated learning to address it.In this framework, a cross-border dataset is produced first, and then five main modules are designed.Then, a neural network is used as the base model for federated training to train a global model without sending any data outside of the country.Last, experimental evaluations check how well the system works.The results show that the proposed method has an accuracy of 0.892 and a robustness of 0.912, meaning it works well even while processing many types of data and protecting privacy.This is far better than standard methods.This technology offers a safe and effective way to work together on data across borders.