P300-based deception detection in simulated network fraud condition

A deception detection experiment with three-stimulus guilty knowledge test paradigm in simulated network fraud condition was conducted. The raw electroencephalography ( EEG) signals were acquired from 12 subjects during the deception detection experiment. Then a multi-domain EEG signal processing method was proposed, preprocessing the raw EEG signals and extracting features in temporal, spectral and spatial domains. Subsequently, genetic algorithm was implemented to obtain optimal feature subset and linear discriminant analysis was used for classification. Experiment results obtained by the proposed technique confirmed its effectiveness in deception detection under simulated network fraud condition.