Nonlinear System Modeling via Knot-Optimizing B-Spline Networks

K.F.C. Yiu, Song Wang, K.L. Teo, A.C. Tsoi

    Research output: Contribution to journalArticle

    25 Citations (Scopus)

    Abstract

    In using the B-spline network for nonlinear system modeling, owing to a lack of suitable theoretical results, it is quite difficult to choose an appropriate set of knot points to achieve a good network structure for minimizing, say, a minimum error criterion. In this paper, a novel knot-optimizing B-spline network is proposed to approximate general nonlinear system behavior. The knot points are considered to be independent variables in the B-spline network and are optimized together with the B-spline expansion coefficients. A simulated annealing algorithm with an appropriate search strategy is used as an optimization algorithm for the training process in order to avoid any possible local minima. Examples involving dynamic systems up to six dimensions in the input space to the network are solved by the proposed method to illustrate the effectiveness of this approach.
    Original languageEnglish
    Pages (from-to)1013-1022
    JournalIEEE Transactions on Neural Networks
    Volume12
    Issue number5
    Publication statusPublished - 2001

    Fingerprint

    Splines
    Nonlinear systems
    Simulated annealing
    Dynamical systems

    Cite this

    Yiu, K. F. C., Wang, S., Teo, K. L., & Tsoi, A. C. (2001). Nonlinear System Modeling via Knot-Optimizing B-Spline Networks. IEEE Transactions on Neural Networks, 12(5), 1013-1022.
    Yiu, K.F.C. ; Wang, Song ; Teo, K.L. ; Tsoi, A.C. / Nonlinear System Modeling via Knot-Optimizing B-Spline Networks. In: IEEE Transactions on Neural Networks. 2001 ; Vol. 12, No. 5. pp. 1013-1022.
    @article{429f09ddf51244dc85be572ecaba36d2,
    title = "Nonlinear System Modeling via Knot-Optimizing B-Spline Networks",
    abstract = "In using the B-spline network for nonlinear system modeling, owing to a lack of suitable theoretical results, it is quite difficult to choose an appropriate set of knot points to achieve a good network structure for minimizing, say, a minimum error criterion. In this paper, a novel knot-optimizing B-spline network is proposed to approximate general nonlinear system behavior. The knot points are considered to be independent variables in the B-spline network and are optimized together with the B-spline expansion coefficients. A simulated annealing algorithm with an appropriate search strategy is used as an optimization algorithm for the training process in order to avoid any possible local minima. Examples involving dynamic systems up to six dimensions in the input space to the network are solved by the proposed method to illustrate the effectiveness of this approach.",
    author = "K.F.C. Yiu and Song Wang and K.L. Teo and A.C. Tsoi",
    year = "2001",
    language = "English",
    volume = "12",
    pages = "1013--1022",
    journal = "IEEE Transactions on Neural Networks",
    issn = "1045-9227",
    publisher = "IEEE, Institute of Electrical and Electronics Engineers",
    number = "5",

    }

    Yiu, KFC, Wang, S, Teo, KL & Tsoi, AC 2001, 'Nonlinear System Modeling via Knot-Optimizing B-Spline Networks' IEEE Transactions on Neural Networks, vol. 12, no. 5, pp. 1013-1022.

    Nonlinear System Modeling via Knot-Optimizing B-Spline Networks. / Yiu, K.F.C.; Wang, Song; Teo, K.L.; Tsoi, A.C.

    In: IEEE Transactions on Neural Networks, Vol. 12, No. 5, 2001, p. 1013-1022.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - Nonlinear System Modeling via Knot-Optimizing B-Spline Networks

    AU - Yiu, K.F.C.

    AU - Wang, Song

    AU - Teo, K.L.

    AU - Tsoi, A.C.

    PY - 2001

    Y1 - 2001

    N2 - In using the B-spline network for nonlinear system modeling, owing to a lack of suitable theoretical results, it is quite difficult to choose an appropriate set of knot points to achieve a good network structure for minimizing, say, a minimum error criterion. In this paper, a novel knot-optimizing B-spline network is proposed to approximate general nonlinear system behavior. The knot points are considered to be independent variables in the B-spline network and are optimized together with the B-spline expansion coefficients. A simulated annealing algorithm with an appropriate search strategy is used as an optimization algorithm for the training process in order to avoid any possible local minima. Examples involving dynamic systems up to six dimensions in the input space to the network are solved by the proposed method to illustrate the effectiveness of this approach.

    AB - In using the B-spline network for nonlinear system modeling, owing to a lack of suitable theoretical results, it is quite difficult to choose an appropriate set of knot points to achieve a good network structure for minimizing, say, a minimum error criterion. In this paper, a novel knot-optimizing B-spline network is proposed to approximate general nonlinear system behavior. The knot points are considered to be independent variables in the B-spline network and are optimized together with the B-spline expansion coefficients. A simulated annealing algorithm with an appropriate search strategy is used as an optimization algorithm for the training process in order to avoid any possible local minima. Examples involving dynamic systems up to six dimensions in the input space to the network are solved by the proposed method to illustrate the effectiveness of this approach.

    M3 - Article

    VL - 12

    SP - 1013

    EP - 1022

    JO - IEEE Transactions on Neural Networks

    JF - IEEE Transactions on Neural Networks

    SN - 1045-9227

    IS - 5

    ER -