首页> 外国专利> METHOD FOR MANAGING SELF-ADAPTIVE DYNAMIC MULTIDIMENSIONAL ARRAY TO ENABLE NEUROPLASTICITY FOR SELF-ADAPTIVE AND PROGRESSIVE MACHINE LEARNING AND COMPUTER SOFTWARE USING THE SAME

METHOD FOR MANAGING SELF-ADAPTIVE DYNAMIC MULTIDIMENSIONAL ARRAY TO ENABLE NEUROPLASTICITY FOR SELF-ADAPTIVE AND PROGRESSIVE MACHINE LEARNING AND COMPUTER SOFTWARE USING THE SAME

机译:管理自适应多维度动态矩阵以使自适应和渐进式机器学习和计算机软件具有相同的神经可塑性的方法

摘要

The present invention relates to a method for managing a self-adaptive dynamic multidimensional array (saDMA) for autonomous decision making, and computer software and computing entity using the same. Individual cells included in the self-adaptive dynamic multidimensional array are distinguished by parameters including at least individual events, individual states and individual contexts. The individual cells include operations to be performed according to the parameters corresponding to the individual cells. A method according to the present invention includes the steps of: acquiring, by a computing entity, a list of operations suitable for an event if at least one event is obtained, the computing entity based on the self-adaptive dynamic multidimensional array; performing the individual operations belonging to the list of suitable operations; and rearranging the self-adaptive dynamic multidimensional array based on the result of the performance. Thus, the neuroplasticity and self-adaptive flexibility of the human brain can be implemented in computer software and computing entities related to the artificial intelligence.
机译:本发明涉及一种用于管理用于自主决策的自适应动态多维阵列(saDMA)的方法,以及使用该方法的计算机软件和计算实体。自适应动态多维数组中包含的单个单元格通过至少包括单个事件,单个状态和单个上下文的参数进行区分。各个单元包括要根据与各个单元相对应的参数执行的操作。根据本发明的方法包括以下步骤:计算实体基于自适应动态多维阵列,如果获得至少一个事件,则获取适合于事件的操作列表;以及执行属于适当操作列表的各个操作;并根据性能结果重新排列自适应动态多维数组。因此,可以在与人工智能有关的计算机软件和计算实体中实现人脑的神经可塑性和自适应灵活性。

著录项

  • 公开/公告号KR20190001711A

    专利类型

  • 公开/公告日2019-01-07

    原文格式PDF

  • 申请/专利权人 BAEK OCK KEE;

    申请/专利号KR20170081667

  • 发明设计人 BAEK OCK KEE;

    申请日2017-06-28

  • 分类号G06N3/04;G06F9/44;G06N3/08;

  • 国家 KR

  • 入库时间 2022-08-21 11:51:51

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