基于模型融合的案件知情者识别方法

本发明公开了一种基于模型融合的案件知情者识别方法，包括以下步骤，提取各个被测试者在观看单一图片时的32维眼动特征；基于32维眼动特征训练支撑向量机模型A，来识别每个被测试者在单一图片时的言语置信度，并输出每个被测试者在单一图片时的概率f1(xi)和f2(xi)；提取各个被测试者在观看组合图片时的110维眼动特征；基于110维眼动特征训练支撑向量机模型B，来识别每个被测试者在组合图片时的言语置信度，并输出每个被测试者在组合图片时的概率g1(xi)和g2(xi)；运用乘法规则，融合支撑向量机模型A和B的分类器概率，得到联合概率，取各个被测试者的概率最大的类别为最后的决策结果。本发明可以有效抑制反测...

Full description

Saved in:
Bibliographic Details
Format Patent
LanguageChinese
Published 09.04.2021
Subjects
CALCULATING
COMPUTING
COUNTING
DIAGNOSIS
HANDLING RECORD CARRIERS
HUMAN NECESSITIES
HYGIENE
IDENTIFICATION
MEDICAL OR VETERINARY SCIENCE
PHYSICS
PRESENTATION OF DATA
RECOGNITION OF DATA
RECORD CARRIERS
SURGERY
Online AccessGet full text

Cover

More Information
Summary:本发明公开了一种基于模型融合的案件知情者识别方法，包括以下步骤，提取各个被测试者在观看单一图片时的32维眼动特征；基于32维眼动特征训练支撑向量机模型A，来识别每个被测试者在单一图片时的言语置信度，并输出每个被测试者在单一图片时的概率f1(xi)和f2(xi)；提取各个被测试者在观看组合图片时的110维眼动特征；基于110维眼动特征训练支撑向量机模型B，来识别每个被测试者在组合图片时的言语置信度，并输出每个被测试者在组合图片时的概率g1(xi)和g2(xi)；运用乘法规则，融合支撑向量机模型A和B的分类器概率，得到联合概率，取各个被测试者的概率最大的类别为最后的决策结果。本发明可以有效抑制反测谎手段，提高了算法效率。 The invention discloses a case insider identification method based on model fusion, comprising the following steps of extracting 32-dimensional eye movement features of each subject when viewing a single picture; based on a 32-dimensional eye movement feature train support vector machine model A, recognizing that verbal confidence of each subject in a single picture, and outputting the probabilityf1 (xi) and f2 (xi) of each subject in a single picture; extracting 110-dimensional eye movement features of each subject when viewing a combined picture; training a support vector machine model B based on a 110-dimensional eye movement feature to recognize that verbal confidence level of each subject when com
Bibliography:Application Number: CN201811135018