TY - GEN
T1 - Estimation of the number of sources by using an MMSE-based MDL criterion without eigendecomposition
AU - Huang, Lei
AU - Long, Teng
AU - Mao, Erke
PY - 2009
Y1 - 2009
N2 - A new computationally simple MDL approach is addressed in this paper. Unlike the eigenvalue-based MDL methods, the proposed method suggests to use the minimum mean square errors (MMSEs) of the multi-stage Wiener filter (MSWF) to calculate the description length required to encode the observed data. As a result, the proposed method is more robust to the nonuniform noise than the eigenvalue-based MDL methods. On the other hand, since the proposed method does not involve the estimation of the observed covariance matrix and its eigedecomposition, its computational complexity can be significantly reduced. Numerical results are presented to illustrate the consistency and robustness of the proposed method.
AB - A new computationally simple MDL approach is addressed in this paper. Unlike the eigenvalue-based MDL methods, the proposed method suggests to use the minimum mean square errors (MMSEs) of the multi-stage Wiener filter (MSWF) to calculate the description length required to encode the observed data. As a result, the proposed method is more robust to the nonuniform noise than the eigenvalue-based MDL methods. On the other hand, since the proposed method does not involve the estimation of the observed covariance matrix and its eigedecomposition, its computational complexity can be significantly reduced. Numerical results are presented to illustrate the consistency and robustness of the proposed method.
UR - http://www.scopus.com/inward/record.url?scp=70449669299&partnerID=8YFLogxK
U2 - 10.1109/ICSPS.2009.27
DO - 10.1109/ICSPS.2009.27
M3 - Conference contribution
AN - SCOPUS:70449669299
SN - 9780769536545
T3 - 2009 International Conference on Signal Processing Systems, ICSPS 2009
SP - 8
EP - 12
BT - 2009 International Conference on Signal Processing Systems, ICSPS 2009
T2 - 2009 International Conference on Signal Processing Systems, ICSPS 2009
Y2 - 15 May 2009 through 17 May 2009
ER -