【24h】

ROC Curves for Multivariate Biometric Matching Models

机译:用于多变量生物识别匹配模型的ROC曲线

获取原文

摘要

The biometric matching problem is a two (within or between) class classification problem where two types of errors (FRR and FAR) are introduced. While the receiver operating characteristic or ROC curve, which is a plot of FRR and FAR, can be easily obtained in the simple matching model, it is non-trivial to obtain a curve in the multivariate matching model. Here the problem of obtaining the ROC curves for several parametric and nonparametric matchers in a multidimensional case is considered. Graphical decision boundaries are also given for each matcher.
机译:生物识别匹配问题是介绍了两种类型的类分类问题,其中引入了两种类型的错误(FRR和FAR)。虽然在简单匹配模型中可以容易地获得作为FRR和远的曲线图的接收器操作特性或ROC曲线,但是在多变量匹配模型中获得曲线是不普遍的。这里考虑在多维案例中获取多个参数和非参数匹配器的ROC曲线的问题。每个匹配器也给出了图形决策边界。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

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