Information criteria determine the number of active sources

T.R. Knosche, T.R. Knosche, E.M. Berends, M.J. Peters, Bert Jagers

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    42 Downloads (Pure)

    Abstract

    With the neuroelectromagnetic inverse problem, the optimal choice of the number of sources is a difficult problem, especially in the presence of correlated noise. In this paper we present a number of information criteria that help to solve this problem. They are based on the probability density function of the measurements or their eigenvalues. Make use of the Akaike or MDL (minimum description length) correction term and all employ some sort of noise information. By extensive simulations we investigated the conditions under which these criteria yield reliable estimations. We were able to quantify two major factors of influence: (1) the precision of the noise information and (2) the signal-to-noise ratio (SNR). Here defined as the ratio of the smallest signal eigenvalues and the average of the noise eigenvalues. Furthermore, we found that the Akaike correction term tends to overestimate, due to its greater sensibility to the precision of the noise information
    Original languageUndefined
    Title of host publicationProc. of the IBME
    Place of PublicationArnhem
    PublisherIEEE
    Pages83-87
    Number of pages5
    ISBN (Print)90-365-0845-2
    DOIs
    Publication statusPublished - 30 Sep 1996
    Event18th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 1996: Bridging Disciplines for Biomedicine - Amsterdam, Netherlands
    Duration: 31 Oct 19963 Nov 1996
    Conference number: 18

    Publication series

    Name
    PublisherIEEE
    Volume2

    Conference

    Conference18th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 1996
    Abbreviated titleEMBC
    CountryNetherlands
    CityAmsterdam
    Period31/10/963/11/96

    Keywords

    • IR-56014
    • METIS-130341

    Cite this

    Knosche, T. R., Knosche, T. R., Berends, E. M., Peters, M. J., & Jagers, B. (1996). Information criteria determine the number of active sources. In Proc. of the IBME (pp. 83-87). Arnhem: IEEE. https://doi.org/10.1109/IEMBS.1996.651992
    Knosche, T.R. ; Knosche, T.R. ; Berends, E.M. ; Peters, M.J. ; Jagers, Bert. / Information criteria determine the number of active sources. Proc. of the IBME. Arnhem : IEEE, 1996. pp. 83-87
    @inproceedings{99fa35367d48415d80ecc9bfc440734d,
    title = "Information criteria determine the number of active sources",
    abstract = "With the neuroelectromagnetic inverse problem, the optimal choice of the number of sources is a difficult problem, especially in the presence of correlated noise. In this paper we present a number of information criteria that help to solve this problem. They are based on the probability density function of the measurements or their eigenvalues. Make use of the Akaike or MDL (minimum description length) correction term and all employ some sort of noise information. By extensive simulations we investigated the conditions under which these criteria yield reliable estimations. We were able to quantify two major factors of influence: (1) the precision of the noise information and (2) the signal-to-noise ratio (SNR). Here defined as the ratio of the smallest signal eigenvalues and the average of the noise eigenvalues. Furthermore, we found that the Akaike correction term tends to overestimate, due to its greater sensibility to the precision of the noise information",
    keywords = "IR-56014, METIS-130341",
    author = "T.R. Knosche and T.R. Knosche and E.M. Berends and M.J. Peters and Bert Jagers",
    year = "1996",
    month = "9",
    day = "30",
    doi = "10.1109/IEMBS.1996.651992",
    language = "Undefined",
    isbn = "90-365-0845-2",
    publisher = "IEEE",
    pages = "83--87",
    booktitle = "Proc. of the IBME",
    address = "United States",

    }

    Knosche, TR, Knosche, TR, Berends, EM, Peters, MJ & Jagers, B 1996, Information criteria determine the number of active sources. in Proc. of the IBME. IEEE, Arnhem, pp. 83-87, 18th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 1996, Amsterdam, Netherlands, 31/10/96. https://doi.org/10.1109/IEMBS.1996.651992

    Information criteria determine the number of active sources. / Knosche, T.R.; Knosche, T.R.; Berends, E.M.; Peters, M.J.; Jagers, Bert.

    Proc. of the IBME. Arnhem : IEEE, 1996. p. 83-87.

    Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

    TY - GEN

    T1 - Information criteria determine the number of active sources

    AU - Knosche, T.R.

    AU - Knosche, T.R.

    AU - Berends, E.M.

    AU - Peters, M.J.

    AU - Jagers, Bert

    PY - 1996/9/30

    Y1 - 1996/9/30

    N2 - With the neuroelectromagnetic inverse problem, the optimal choice of the number of sources is a difficult problem, especially in the presence of correlated noise. In this paper we present a number of information criteria that help to solve this problem. They are based on the probability density function of the measurements or their eigenvalues. Make use of the Akaike or MDL (minimum description length) correction term and all employ some sort of noise information. By extensive simulations we investigated the conditions under which these criteria yield reliable estimations. We were able to quantify two major factors of influence: (1) the precision of the noise information and (2) the signal-to-noise ratio (SNR). Here defined as the ratio of the smallest signal eigenvalues and the average of the noise eigenvalues. Furthermore, we found that the Akaike correction term tends to overestimate, due to its greater sensibility to the precision of the noise information

    AB - With the neuroelectromagnetic inverse problem, the optimal choice of the number of sources is a difficult problem, especially in the presence of correlated noise. In this paper we present a number of information criteria that help to solve this problem. They are based on the probability density function of the measurements or their eigenvalues. Make use of the Akaike or MDL (minimum description length) correction term and all employ some sort of noise information. By extensive simulations we investigated the conditions under which these criteria yield reliable estimations. We were able to quantify two major factors of influence: (1) the precision of the noise information and (2) the signal-to-noise ratio (SNR). Here defined as the ratio of the smallest signal eigenvalues and the average of the noise eigenvalues. Furthermore, we found that the Akaike correction term tends to overestimate, due to its greater sensibility to the precision of the noise information

    KW - IR-56014

    KW - METIS-130341

    U2 - 10.1109/IEMBS.1996.651992

    DO - 10.1109/IEMBS.1996.651992

    M3 - Conference contribution

    SN - 90-365-0845-2

    SP - 83

    EP - 87

    BT - Proc. of the IBME

    PB - IEEE

    CY - Arnhem

    ER -

    Knosche TR, Knosche TR, Berends EM, Peters MJ, Jagers B. Information criteria determine the number of active sources. In Proc. of the IBME. Arnhem: IEEE. 1996. p. 83-87 https://doi.org/10.1109/IEMBS.1996.651992