METHOD AND APPARATUS FOR PROVIDING CUSTOMIZED ARTIFICIAL NEURAL NETWORK MODEL BY USING FEDERATED LEARNING

The present invention relates to a data market using federated learning. An operation method of a central server according to an embodiment comprises the steps of: providing data provider information to a consumer; receiving data request information comprising an artificial neural network model from...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: KIM, Jong Yul, PARK, Wei Jin, KOH, Eui Yul, CHOI, Jong Myeong, SHIN, Hyun Kyeong, LEE, Su Gi
Format: Patent
Sprache:eng ; fre ; kor
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The present invention relates to a data market using federated learning. An operation method of a central server according to an embodiment comprises the steps of: providing data provider information to a consumer; receiving data request information comprising an artificial neural network model from the consumer; transmitting the data request information to a target data provider; receiving, from the target data provider, the result of learning by an artificial neural network model that has been individually trained; generating a common model on the basis of the result of learning; and transmitting the common model to the consumer. La présente invention concerne un marché de données utilisant l'apprentissage fédéré. Un procédé de fonctionnement d'un serveur central selon un mode de réalisation comprend les étapes consistant à : fournir des informations de fournisseur de données à un consommateur ; recevoir, du consommateur, des informations de demande de données comprenant un modèle de réseau neuronal artificiel ; transmettre les informations de demande de données à un fournisseur de données cible ; recevoir, du fournisseur de données cible, le résultat d'apprentissage par un modèle de réseau neuronal artificiel qui a été entraîné individuellement ; générer un modèle commun sur la base du résultat d'apprentissage ; et transmettre le modèle commun au consommateur. 연합 학습을 이용한 데이터 마켓에 관한 것이다. 일 실시예에 따른 중앙 서버의 동작 방법은 수요자에게 데이터 공급자 정보를 제공하는 단계, 수요자로부터 인공 신경망 모델을 포함하는 데이터 요청 정보를 수신하는 단계, 데이터 요청 정보를 타겟 데이터 공급자에게 전달하는 단계, 타겟 데이터 공급자로부터, 개별적으로 학습된 인공 신경망 모델의 학습 결과를 수신하는 단계, 학습 결과에 기초하여, 공통 모델을 생성하는 단계 및 공통 모델을 수요자에게 전달하는 단계를 포함한다.