• DocumentCode
    2762971
  • Title

    Classification of Old/New effects during memory retrieval Using Committee Machine: an Event-Related Potential study

  • Author

    Rezania, S. ; Nasrabadi, A.M. ; Abootalebi, V.

  • Author_Institution
    Res. Center of Intell. Signal Process., Tehran
  • fYear
    2008
  • fDate
    18-20 Dec. 2008
  • Firstpage
    1
  • Lastpage
    4
  • Abstract
    The task of old/new items recognition associated with memory process has assessed by ERPs in present study. The recognition paradigm with random shape items was administered to 27 young participants. Event-related potentials were recorded during an old/new discrimination task on Fz, Cz and Pz channels. In order to classify old/new responses, wavelet features which are compatible with ERP characteristics in Time-frequency domain were used as discriminator features. Best features were selected using t-test and stepwise linear discriminant analysis method afterward. Linear discriminant analysis was applied as classifier between two groups. Finally, using committee machine, event related potentials associated with old/new effects, were separated with 96.4% accuracy in test group.
  • Keywords
    bioelectric potentials; neurophysiology; time-frequency analysis; committee machine; event related potential; linear discriminant analysis; memory retrieval; random shape items; time frequency domain; Biomedical engineering; Electrodes; Enterprise resource planning; Feature extraction; Information retrieval; Linear discriminant analysis; Protocols; Testing; Time frequency analysis; Wavelet domain; Episodic memory; Event-Related Potentials; Linear discriminant analysis; Stepwise linear discriminant analysis; T-test; Time-frequency domain; wavelet;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Biomedical Engineering Conference, 2008. CIBEC 2008. Cairo International
  • Conference_Location
    Cairo
  • Print_ISBN
    978-1-4244-2694-2
  • Electronic_ISBN
    978-1-4244-2695-9
  • Type

    conf

  • DOI
    10.1109/CIBEC.2008.4786066
  • Filename
    4786066