S. Miyamoto

University of Tsukuba, Tsukuba, Ibaraki, Japan

Are you S. Miyamoto?

Claim your profile

Publications (3)0 Total impact

  • K. Umayahara · S. Miyamoto · Y. Nakamori
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper considers the problem of detecting local linear substructures of a system in a high-dimensional data space by applying a fuzzy clustering technique. We propose a linear fuzzy clustering method using eigenvalues of the fuzzy scatter matrix in the objective function for optimizing the dimensional coefficients. The optimal solutions for the objective function and some illustrative examples are shown in this paper
    IFSA World Congress and 20th NAFIPS International Conference, 2001. Joint 9th; 08/2001
  • O. Takata · S. Miyamoto · K. Umayahara
    [Show abstract] [Hide abstract]
    ABSTRACT: Fuzzy c-means is well-known among the various methods of fuzzy cluster analysis. L<sub>1</sub>-based fuzzy c-means has also been studied in recent years. This paper discusses the L<sub>1</sub>-based fuzzy c-means of data with fuzzy uncertainties. The data unit is supposed to be the Cartesian product of fuzzy numbers. The metric between a data unit with uncertainty and a cluster center is defined using minimum and maximum distances. The fuzzy c-means algorithm is an alternative procedure for the optimization of the cluster center and the fuzzy set membership, while the solution of the cluster center for uncertain data cannot be obtained directly. An algorithm for the solution of cluster centers based on the L<sub>1</sub> metric for uncertain data is developed in this paper. Using this algorithm, an exact alternate optimization procedure is obtained. Numerical examples show that the results for uncertain data are different from the results for data without uncertainties
    IFSA World Congress and 20th NAFIPS International Conference, 2001. Joint 9th; 08/2001
  • S. Miyamoto · K. Umayahara
    [Show abstract] [Hide abstract]
    ABSTRACT: A quadratic regularization method is proposed as a variation of the fuzzy c-means. The standard fuzzy c-means is regarded as a regularization of the crisp k-means, and hence other regularization methods can be studied as fuzzy versions of the crisp c-means. A new algorithm for calculating membership values is derived, whereas calculation of cluster centers is similar to the standard method. The nearest prototype classification functions which has been derived from the standard fuzzy c-means is transformed into the corresponding method within the present method of quadratic regularization. It should be noted that the present method yields piecewise linear classification functions
    Fuzzy Systems Proceedings, 1998. IEEE World Congress on Computational Intelligence., The 1998 IEEE International Conference on; 06/1998