視覺誘發(fā)腦電信號的少次提取.pdf

1、河北工業(yè)大學碩士學位論文視覺誘發(fā)腦電信號的少次提取姓名：劉建輝申請學位級別：碩士專業(yè)：計算機應(yīng)用技術(shù)指導教師：吳清20071101視覺誘發(fā)腦電信號的少次提取iiTHESINGLETRIALESTIMATIONOfVISUALEVOKEDPOTENTIALABSTRACTVisualevokedpotentials(VEP)isatuftyelectricitysignalpotentialofthecerebrumctexinrespo

2、nsetothestimulus.FromthepointofviewofclinicaldiagnosistheextractionofVEPhasveryimptantsignificance.ThesingletrialestimationofVEPhasbecometheresearchobjectiveofmanypepole.Manypepolehavebeenputfwardmanymethodstoresolvethep

3、roblemofVEPextraction.Theindependentcomponentanalysis(ICA)isanewdevelopedstatisticalmethodinrecentyears.Thismethodisbasedonhighderstatistics.ICAistoseparatetheiginalsignalsintoseveralindependentcomponentsbyingaparticular

4、criterionanoptimalalgithm.Thoughthespecificinfmationofthesourcesmixingsystemareunknowntheprobabilityfthesourcedecompositioncanberealizedbyusingtheprincipleofstatisticalindependence.InthispaperthebasicknowledgeofICAhasbee

5、nfirstintroducedincludingthestatisticstheyinfmationtheythatarenecessarytounderstingmasteringICAtechnique.ThenthebasicprincipleofICAhasbeenintroduced.Thispaperintroducedthefamousindependentcomponentanalysisalgithm:twocate

6、giesofindependentcomponentanalysisalgithmwhichbasedoninfmationtheystatisticalkurtosisthey.AnotherresearchfocusofthispaperisFastICAalgithm.Accdingtoindependentcomponentanalysismathematicalbasisofthetheythispaperpointedout

7、theinadequateofNewtonsmethodintheindependentcomponentanalysisraiseditssolution:ApplyingsimplifiedNewtonmethodtoreducetheamountofcomputationAddingadditionalconstraintstoensuretheconvergencefunction.Applyingtheimprovedalgi

8、thmAsimulationexperimentwasdone.Theresultsshowthattheimprovedalgithmcanreducedthenumberofiterationsprovedthefeasibilityofimprovingthealgithm.Attheendofthispaperwegiveconclusionsintroduceourfuturewks.KEYWDS:visualevokedpo