首页> 外国专利> Simulation neural process for Hidden Markov models of signal processing, involves enabling allocation of input sample to output sample by neural synapses simulation that is performed by differential tables

Simulation neural process for Hidden Markov models of signal processing, involves enabling allocation of input sample to output sample by neural synapses simulation that is performed by differential tables

机译:用于信号处理的隐马尔可夫模型的模拟神经过程,涉及通过差分表执行的神经突触模拟实现输入样本到输出样本的分配

摘要

The method involves providing a differential table for 8 bits, 256 memory cells, and for 16 bits, 65536 memory cells, which contains counter notes loaded with different signal sources and allows further low-level differential tables. Allocation of an input sample to an output sample is enabled by neural synapses simulation that is performed by the differential tables.
机译:该方法涉及为8位256个存储单元提供差分表,为16位65536个存储单元提供差分表,其中包含加载有不同信号源的计数器注释,并允许进一步的低级差分表。通过差分表执行的神经突触模拟,可以将输入样本分配到输出样本。

著录项

  • 公开/公告号DE102007018976A1

    专利类型

  • 公开/公告日2008-10-23

    原文格式PDF

  • 申请/专利权人 SECNJAK RICHARD;

    申请/专利号DE20071018976

  • 发明设计人

    申请日2007-04-21

  • 分类号G06N3/00;

  • 国家 DE

  • 入库时间 2022-08-21 19:49:15

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