应用图象识别法预测强震危险区存在的问题及讨论
DOI:
作者:
作者单位:

作者简介:

通讯作者:

基金项目:


PROBLEMS AND DISCUSSIONS ON THE APPLICATION OF PATTERN RECOGNITION IN ZONING STRONG MOTION AREA
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
    摘要:

    应用图象识别法预测强震危险区虽然国内外都有报道,但也还存在一些问题,近来国内已有人发现[1],用图象识别法确定地震危险区存在一些缺陷,它直接影响到结果的可靠程度。为了进一步从理论上搞清产生这些问题的根源及其规律,下面我们就两类图象识别问题的数学模型,根据线性不等式组的理论进行严格的数学讨论。

    Abstract:

    There are two problems in zoning the strong motion area by means of pattern recognition.One is how to decide whether the training set is divi-sible linearly,which is directly related to the problem of convergence of the iteration process.The other problem is the conflicting results of pred-iction that come from the differences of recognition functions which are brought about by the fact that the solution of training is not unique.In this paper these two problems are discussed according to the theory of system of linearly inequalities and a preliminary result has been obtain-ed.

    参考文献
    相似文献
    引证文献
引用本文

毛卫民.应用图象识别法预测强震危险区存在的问题及讨论[J].地震工程学报,1982,4(4):120-123. Mao Weimin. PROBLEMS AND DISCUSSIONS ON THE APPLICATION OF PATTERN RECOGNITION IN ZONING STRONG MOTION AREA[J]. China Earthquake Engineering Journal,1982,4(4):120-123.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
历史
  • 收稿日期:1981-11-21
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2017-07-04