Method for training an algorithm and method for providing a service based on the trained algorithm
A computer implemented method for training algorithms, comprising the steps of: providing at least one data owner (102-207) having datasets, at least one algorithm provider (101) having algorithms to be trained and a blockchain network (150) in which a training management provider (160) is deployed;...
Gespeichert in:
Hauptverfasser: | , , , , , , , , |
---|---|
Format: | Patent |
Sprache: | eng ; heb |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | A computer implemented method for training algorithms, comprising the steps of: providing at least one data owner (102-207) having datasets, at least one algorithm provider (101) having algorithms to be trained and a blockchain network (150) in which a training management provider (160) is deployed; sending, by an algorithm provider (101), a training request (201) to the training management provider (160), including a training manifest that determines execution features of the training of the algorithm to be trained and minimum requirements to be met by the training; sending the training manifest to the at least one data owner (102-207); sending by each data owner (102, 104, 105, 107) willing to participating in training the algorithm, an acceptance including an acceptance manifest comprising means ensuring the identity of the participating data owner; sending (204), by the training management provider (160) and once the minimal requirements have been met, the acceptance manifests to the algorithm provider (101); generating, by the algorithm provider (101), a compiled multi-party computation code (208) based on the received acceptance manifests; sending (300) the algorithm to be trained and the compiled multi-party computation code (208) to the one or more data owners participating in training the algorithm; training, by each participating data owner the received algorithm with its corresponding datasets creating a local model; executing the compiled multi-party computation code generating a multi-party peer-to-peer network (200); and generating by means of the multi-party peer-to-peer network (200), a model (711) based on the local models and the algorithm to be trained. A computer implemented method for providing a service based on the optimized model obtained in the method previously mentioned. |
---|