首页> 外国专利> A system of pretreatment and storage of heterogenous big-data for deep-learning of big-data

A system of pretreatment and storage of heterogenous big-data for deep-learning of big-data

机译:用于深度学习大数据的异构大数据预处理和存储系统

摘要

Deep learning of heterogeneous big data, which collects heterogeneous big data generated from various sensors at the work site, but stores atypical heterogeneous big data in a format having a uniform structure by pre-processing it so that all can be applied directly to deep learning learning A system for preprocessing and storing heterogeneous big data for learning, comprising: a data collection unit for collecting heterogeneous data from various sensors; A preprocessor for generating preprocessed data by converting the collected data according to the preprocessing method of each heterogeneous data for each heterogeneous data; A data storage unit for storing the pre-processed data according to a predetermined structure; And a meta information insertion unit generating and inserting meta data for the pre-processed data. By pre-processing and storing heterogeneous big data collected in real time in a uniform structure by the system as described above, it is possible to eliminate repetitive pre-processing tasks when deep learning learning is continuously performed, thereby improving learning efficiency. have.
机译:异构大数据的深度学习,它收集工作现场各种传感器生成的异构大数据,但通过预处理将非典型异构大数据存储为具有统一结构的格式,以便将其直接应用于深度学习一种用于预处理和存储用于学习的异构大数据的系统,包括:数据收集单元,用于从各种传感器收集异构数据;以及一种预处理器,用于通过根据每个异构数据的每个异构数据的预处理方法转换收集到的数据来生成预处理数据;数据存储单元,用于根据预定结构存储预处理后的数据;并且元信息插入单元生成并插入用于预处理数据的元数据。通过如上所述的系统,以统一的结构对实时收集的异构大数据进行预处理和存储,可以消除连续进行深度学习时重复的预处理任务,从而提高学习效率。有。

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