专利名称：DISTRIBUTED MACHINE LEARNING
SYSTEMS, APPARATUS, AND METHODS
发明人：SZETO, Christopher,BENZ, Stephen,
Charles,WITCHEY, Nicholas, J.
申请号：EP17831625.3申请日：20170717公开号：EP3485436A1公开日：20190522
摘要：A distributed, online machine learning system is presented. Contemplatedsystems include many private data servers, each having local private data. Researcherscan request that relevant private data servers train implementations of machine learningalgorithms on their local private data without requiring de-identification of the privatedata or without exposing the private data to unauthorized computing systems. Theprivate data servers also generate synthetic or proxy data according to the datadistributions of the actual data. The servers then use the proxy data to train proxymodels. When the proxy models are sufficiently similar to the trained actual models, theproxy data, proxy model parameters, or other learned knowledge can be transmitted toone or more non-private computing devices. The learned knowledge from many privatedata servers can then be aggregated into one or more trained global models withoutexposing private data.
申请人：Nantomics, LLC,Nant Holdings IP, LLC
地址：9920 Jefferson Boulevard Culver City, CA 90232 US,9920 Jefferson BoulevardCulver City, CA 90232 US
国籍：US,US
代理机构：AWA Sweden AB
更多信息请下载全文后查看
因篇幅问题不能全部显示，请点此查看更多更全内容