For Heads-of · Practitioner
Federated Learning
A machine learning approach where models are trained across decentralized data sources without centralizing sensitive data in one location.
- federated learning
- distributed
- privacy
- training
Process
Instead of sending data to a central server, the model is sent to data holders, trained locally, and only model updates are returned to aggregate. The central entity never sees raw data.
Privacy Benefit
Sensitive data never leaves its original location, reducing privacy risk and enabling training on regulated data (health records, financial information) that cannot be centralized.
Challenges
Federated learning is technically complex and computationally expensive. Privacy guarantees still require differential privacy. Communication costs are high.