Back
Federated Learning: Collaborative Machine Learning without Centralized Training
webai.googleblog.com·ai.googleblog.com/2017/04/federated-learning-collaborativ...
This Google AI Blog post introduced federated learning to a broad audience; it is relevant to AI safety discussions around data privacy, decentralized AI governance, and reducing risks of large-scale data centralization by powerful AI developers.
Metadata
Importance: 62/100blog postprimary source
Summary
Google introduces federated learning, a technique that trains machine learning models across many decentralized devices (like smartphones) without centralizing raw user data. Instead of sending data to a server, the model is sent to each device, trained locally, and only model updates (gradients) are aggregated centrally. This approach offers privacy benefits by keeping sensitive user data on-device while still enabling powerful shared models.
Key Points
- •Federated learning trains models on distributed devices by sharing model updates rather than raw data, preserving user privacy.
- •Developed primarily for mobile devices (e.g., Gboard), where personal data is sensitive and bandwidth for uploading raw data is limited.
- •Aggregation of model updates (via Federated Averaging) allows a global model to improve without any single party seeing individual user data.
- •Raises important questions about differential privacy, secure aggregation, and whether gradients themselves can leak private information.
- •Represents a shift in ML infrastructure that has significant implications for privacy-preserving AI deployment and data governance.
Cited by 1 page
| Page | Type | Quality |
|---|---|---|
| AI-Driven Concentration of Power | Risk | 65.0 |
Cached Content Preview
HTTP 200Fetched Mar 20, 202613 KB
# Federated Learning: Collaborative Machine Learning without Centralized Training Data
April 6, 2017
Posted by Brendan McMahan and Daniel Ramage, Research Scientists
## Quick links
- Share
- [Share on Twitter](https://twitter.com/intent/tweet?text=https%3A//research.google/blog/federated-learning-collaborative-machine-learning-without-centralized-training-data/ "Share on Twitter")
- [Share on Facebook](https://www.facebook.com/sharer/sharer.php?u=https%3A//research.google/blog/federated-learning-collaborative-machine-learning-without-centralized-training-data/ "Share on Facebook")
- [Share on LinkedIn](https://www.linkedin.com/shareArticle?url=https%3A//research.google/blog/federated-learning-collaborative-machine-learning-without-centralized-training-data/&mini=true "Share on LinkedIn")
- [Send via Email](mailto:name@example.com?subject=Check%20out%20this%20site&body=Check%20out%20https%3A//research.google/blog/federated-learning-collaborative-machine-learning-without-centralized-training-data/ "Send via Email")
- Copy link
×
Standard machine learning approaches require centralizing the training data on one machine or in a datacenter. And Google has built one of the most secure and robust cloud infrastructures for processing this data to make our services better. Now for models trained from user interaction with mobile devices, we're introducing an additional approach: _Federated Learning_.
Federated Learning enables mobile phones to collaboratively learn a shared prediction model while keeping all the training data on device, decoupling the ability to do machine learning from the need to store the data in the cloud. This goes beyond the use of local models that make predictions on mobile devices (like the [Mobile Vision API](https://developers.google.com/vision/) and [On-Device Smart Reply](https://research.googleblog.com/2017/02/on-device-machine-intelligence.html)) by bringing model _training_ to the device as well.
It works like this: your device downloads the current model, improves it by learning from data on your phone, and then summarizes the changes as a small focused update. Only this update to the model is sent to the cloud, using encrypted communication, where it is immediately averaged with other user updates to improve the shared model. All the training data remains on your device, and no individual updates are stored in the cloud.
| |
| --- |
| [](https://1.bp.blogspot.com/-K65Ed68KGXk/WOa9jaRWC6I/AAAAAAAABsM/gglycD_anuQSp-i67fxER1FOlVTulvV2gCLcB/s1600/FederatedLearning_FinalFiles_Flow%2BChart1.png) |
| Your phone personalizes the model locally, based on your usage (A). Many users' updates
... (truncated, 13 KB total)Resource ID:
a47933706c3362a7 | Stable ID: NWNjODM0ND