Statistical Analysis Of Event-related Potentials In Cognitive Neuroscience Research

The research on the brain mechanism of social decision-making behavior is a hotspot issue in the research of cognitive neural mechanism.The study of the dy-namic process of social decision-making plays an important role in understanding the cognitive and neural mechanisms of social decision-making,especially can im-prove the accuracy of prediction.Event related potential(ERP)is a powerful tool to study the dynamic process of social decision making,which has the character-istics of high temporal resolution and noninvasive.Event related potential(ERP)technology is a special brain evoked potential,which is the response of the brain to the information contained in the stimulus.It is of great theoretical and prac-tical significance to study event-related potentials.The key problem in this field is the statistical inference of the time(starting point)and the amplitude of these components in the decision making process.Functional data analysis is a statistical methods with strong applicability and widely used in recent years,this paper intends to study the ERP data generated from the study of social decision-making dynamics from the perspective of functional data.Cluster using functional data clustering methods under different experimental conditions of frontal and occipital 30 electrodes produced by ERP,whether the po-tential changes of experimental conditions induced distinguishable between frontal and occipital.And use functional data discriminant analysis method to achieve ER-P classification based on less test.The simulation results show that in addition to the electrode f2,fc2,fz,cb2,ft8 the correct rate of other electrodes can reach more than 80%.In the study of prediction problem,we can reduce the testing time,solve the problem of test subjects of fatigue for a long time,but also can greatly reduce the test cost.