首页> 外国专利> MACHINE LEARNING WORKER NODE ARCHITECTURE

MACHINE LEARNING WORKER NODE ARCHITECTURE

机译:机器学习工作者节点架构

摘要

A database contains a corpus of incident reports, a machine learning (ML)model trainedto calculate paragraph vectors of the incident reports, and a look-up settable that contains a listof paragraph vectors respectively associated with sets of the incidentreports. A plurality of MLworker nodes each store the look-up set table and are configured to executethe ML model. Anupdate thread is configured to: determine that the look-up set table hasexpired; update thelook-up set table by: (i) adding a first set of incident reports receivedsince a most recent update of thelook-up set table, and (ii) removing a second set of incident reportscontaining timestamps thatare no longer within a sliding time window; store, in the database, the look-up set table asupdated; and transmit, to the ML worker nodes, respective indications that thelook-up set tablehas been updated.
机译:数据库包含事件报告的语料库,机器学习(ML)模型训练计算事件报告的段落矢量和查询集包含列表的表分别与事件集相关的段落向量的集合报告。多个ML每个工作节点都存储查找集表,并配置为执行ML模型。一个更新线程配置为:确定查找集表具有过期;更新查找设置表,方法是:(i)添加收到的第一组事件报告自从查找设置表,以及(ii)删除第二组事件报告包含的时间戳记不在滑动时间范围内;在数据库中存储外观-设置表为更新;并将相应的指示发送到ML工作节点查找表已经升级。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号