Fisher判别函数在足迹分析性别中的应用
Fisher Discriminant Function in Footprint Analysis of Gender
王东 1张建功 1汤澄清1
作者信息
- 1. 中国刑事警察学院, 辽宁 沈阳 110854
- 折叠
摘要
结合脚型性别差异研究,为了选取出足迹上与性别显著相关的特征,以便通过Fisher判别法建立函数模型,本研究测量了共 7 个与性别显著相关的几何特征,通过T检验和Fisher判别法建立了模型;利用Fisher判别分析建立了典则判别函数和Fisher线性判别函数.实验结果显示,选取的特征对模型影响显著,函数模型预测准确率为 92.3%.该判别函数模型可以显著提高足迹分析性别的准确率,可为技术侦查提供帮助.
Abstract
Combining with the study of gender differences in foot shape,in order to select features significantly correlated with gender on footprints and establish a function model through Fisher discriminant method,this study measures a total of 7 geometric features significantly correlated with gender,and establishes the model through T-test and Fisher discriminant method;a canonical discriminant function and a Fisher linear discriminant function are established using Fisher discriminant analysis.The experimental results show that the selected features have a significant impact on the model,and the prediction accuracy of the function model is 92.3%.This discriminant function model can significantly improve the accuracy of gender footprint analysis and provide assistance for technical investigation.
关键词
足迹/几何特征/Fisher判别分析/性别预测Key words
Footprint/geometric feature/Fisher discriminant analysis/gender prediction引用本文复制引用
出版年
2024
NETL
NSTL
万方数据