首页> 外国专利> NEURAL NETWORK EMBEDDINGS FOR ALARM REPRESENTATION IN DISTRITBUTED NETWORKS

NEURAL NETWORK EMBEDDINGS FOR ALARM REPRESENTATION IN DISTRITBUTED NETWORKS

机译:虚拟网络中报警表示的神经网络嵌入

摘要

A method for training embeddings of network alarms, the network alarms being caused by incidents occurring in a distributed network comprising a plurality of nodes, can include one or more of the following steps: receiving for a given period of time a plurality of network alarms from the plurality of nodes, the plurality of network alarms being caused by a plurality of network incidents occurring in the distributed network; representing the plurality of network alarms into a training set based on co-occurrences of network alarms; and training a plurality of alarm embeddings on the training set using a first neural network model.
机译:一种用于训练网络警报的嵌入方法,该网络警报由包括多个节点的分布式网络中发生的事件引起的,可以包括以下步骤中的一个或多个:从多个网络警报接收给定的时间段。来自多个网络警报多个节点,多个网络警报由分布式网络中发生的多个网络事件引起;基于网络警报的共同发生,将多个网络报警组成为培训集;并使用第一神经网络模型在训练集上训练多个警报嵌入。

著录项

  • 公开/公告号WO2021073707A1

    专利类型

  • 公开/公告日2021-04-22

    原文格式PDF

  • 申请/专利权人 ABOULABAN SAID;

    申请/专利号WO2019EP77693

  • 发明设计人 ABOULABAN SAID;

    申请日2019-10-14

  • 分类号H04L12/24;

  • 国家 EP

  • 入库时间 2022-08-24 18:22:35

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