Abstract
The following quote was obtained by the news editors from the background information supplied by the inventors: “Inner join of records of datasets from multiple data owners may be a prerequisite for subsequent operations for many applications, including the collection of the aggregate statistics, the training of distributed machine learning models, etc. For instance, when one party owns elements’ feature data or attribution data and another party owns the elements’ label or identification, there is a need for a private inner join that does not reveal the intersection before conducting the distributed or federated training in machine learning applications.
NETL
NSTL