摘要
提出一种基于Fisher判别式和事件相关电位(ERP)的心理意识真实性识别新方法,以可识别的有意义的个人信息(如姓名、生日)作为被测试信息,应用隐藏信息测试(CIT)模式对15名受试者进行了测试并记录其脑电信号.将多通道ERP P300幅度和受试者的反应时间组成特征参数,利用Fisher判别式进行心理意识真实性识别分类,获得93.3%的平均识别率,明显高于以往研究方法的识别率.因此,基于多通道ERP和Fisher判别式的方法具有无创性及较高识别率等优点,可以作为一种心理意识真实性检测的新方法.
Fisher discriminant analysis(FDA)and event-related potentials(ERP)are adopted to identify mentality facticity or detect lie.Some identifiably meaning information,such as subject name and birthday,are selected as the concealed information to be identified,15 subjects were tested by concealed information test(CIT)paradigm and their electroence phalographs(EEG) are recorded.Multi-channel ERP P300 amplitude and reaction time of subject are used as the feature parameters and an FDA is applied to identify mentality facticity with a 93.3% of the detection correctness,which is higher than that of methods in the previous researches,and the advantages of non-invasion in the proposed method get obvious.
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2010年第8期132-136,共5页
Journal of Xi'an Jiaotong University
基金
国家自然科学基金资助项目(30870654)
关键词
心理意识
事件相关电位
Fisher判别式
测谎
mentality
event-related potential
Fisher discriminant analysis
lie detection