18283343. METHODS, APPARATUSES, AND SYSTEMS FOR COLLABORATIVELY UPDATING MODEL BY MULTIPLE PARTIES FOR IMPLEMENTING PRIVACY PROTECTION simplified abstract (Alipay (Hangzhou) Information Technology Co., Ltd.)
Contents
- 1 METHODS, APPARATUSES, AND SYSTEMS FOR COLLABORATIVELY UPDATING MODEL BY MULTIPLE PARTIES FOR IMPLEMENTING PRIVACY PROTECTION
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 METHODS, APPARATUSES, AND SYSTEMS FOR COLLABORATIVELY UPDATING MODEL BY MULTIPLE PARTIES FOR IMPLEMENTING PRIVACY PROTECTION - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Potential Applications
- 1.6 Problems Solved
- 1.7 Benefits
- 1.8 Potential Commercial Applications
- 1.9 Possible Prior Art
- 1.10 Original Abstract Submitted
METHODS, APPARATUSES, AND SYSTEMS FOR COLLABORATIVELY UPDATING MODEL BY MULTIPLE PARTIES FOR IMPLEMENTING PRIVACY PROTECTION
Organization Name
Alipay (Hangzhou) Information Technology Co., Ltd.
Inventor(s)
Weiqiang Wang of Hangzhou (CN)
METHODS, APPARATUSES, AND SYSTEMS FOR COLLABORATIVELY UPDATING MODEL BY MULTIPLE PARTIES FOR IMPLEMENTING PRIVACY PROTECTION - A simplified explanation of the abstract
This abstract first appeared for US patent application 18283343 titled 'METHODS, APPARATUSES, AND SYSTEMS FOR COLLABORATIVELY UPDATING MODEL BY MULTIPLE PARTIES FOR IMPLEMENTING PRIVACY PROTECTION
Simplified Explanation
The specification describes a method and system for collaboratively updating a model by multiple parties to implement privacy protection. The system involves a server delivering aggregation results to each participant, who then performs updates on their local model based on common samples and private samples.
- Participants receive aggregation results from the server for each round of common samples.
- Participants update their local model based on the aggregation results and common samples.
- Participants then perform a second update on their model using private samples and sample labels.
- Participants input the next round of common samples into their model for the next iteration.
- The server aggregates the prediction results from all participants for the next round of iteration.
Potential Applications
This technology could be applied in collaborative machine learning projects where privacy protection is crucial, such as healthcare data analysis or financial forecasting.
Problems Solved
This system addresses the challenge of updating a model collaboratively while maintaining privacy and data security among multiple parties.
Benefits
- Enhanced privacy protection in collaborative model updating - Efficient aggregation of prediction results from multiple participants - Facilitates secure data sharing and collaboration in machine learning projects
Potential Commercial Applications
This technology could be valuable in industries such as healthcare, finance, and marketing, where collaborative machine learning projects are common and data privacy is a top priority.
Possible Prior Art
One possible prior art could be the use of federated learning techniques in machine learning, where models are trained across multiple devices without exchanging raw data.
Unanswered Questions
How does this system handle potential conflicts between participants' updates to the model?
The abstract does not provide details on how conflicts in model updates are resolved among participants.
What measures are in place to ensure the security and integrity of the common and private samples used in the model updates?
The abstract does not mention specific security measures or protocols for protecting the common and private samples during the collaborative model updating process.
Original Abstract Submitted
The specification provides a method and a system for collaboratively updating a model by multiple parties for implementing privacy protection. A server can deliver an aggregation result of a t-th round of common samples to each participant i. Each participant i performs first update on a local ith model according to the t-th round of common samples and the aggregation result. Each participant i performs second update on the ith model obtained after the first update based on a first private sample fixed in a local sample set and a sample label thereof. Each participant i inputs a (t+1)th round of common samples that are used for a next round of iteration into the ith model obtained after the second update, and sends an output second prediction result to the server, so the server aggregates n second prediction results corresponding to n participants for a next round of iteration.