Outlier Detecting in Fuzzy Switching Regression Models

  • Hong-bin Shen
  • Jie Yang
  • Shi-tong Wang
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3192)


Fuzzy switching regression models have been extensively used in economics and data mining research. We present a new algorithm named FCWRM (Fuzzy C Weighted Regression Model) to detect the outliers in fuzzy switching regression models while preserving the merits of FCRM algorithm proposed by Hathaway. The theoretic analysis shows that FCWRM can converge to a local minimum of the object function. Several numeric examples demonstrate the effectiveness of algorithm FCWRM.


fuzzy switch regression outlier fuzzy clustering 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Hamermesh, D.S.: Wage bargains, threshold effects, and the Phillips curve. Quart. J. Econ. 84, 501–517 (1970)CrossRefGoogle Scholar
  2. 2.
    Quandt, R.E.: A new approach to estimating switching regressions. J. Amer. Statist. Assoc. 67, 306–310 (1972)zbMATHCrossRefGoogle Scholar
  3. 3.
    Quandt, R.E., Ramsey, J.B.: Estimating mixtures of normal distributions and switching regressions. J. Amer. Statist. Assoc. 73, 730–752 (1978)zbMATHCrossRefMathSciNetGoogle Scholar
  4. 4.
    Hathaway, R., Bezdek, J.: Switching regression models and fuzzy clustering. IEEE Trans. Fuzzy Syst. 1(3), 195–204 (1993)CrossRefGoogle Scholar
  5. 5.
    Caudill, S.B., Acharyal, R.N.: Maximum-likelihood estimation of a mixture of normal regressions: starting values and singularities. Commun. Stat.-Simul. 27(3), 667–674 (1998)zbMATHCrossRefGoogle Scholar
  6. 6.
    Sebert, D.M., Montgomery, D.C., Rollier, D.A.: A Clustering Algorithm for identifying multiple outliers in linear regression. Computational Statistics and Data Analysis 27, 461–484 (1998)zbMATHCrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2004

Authors and Affiliations

  • Hong-bin Shen
    • 1
  • Jie Yang
    • 1
  • Shi-tong Wang
    • 2
  1. 1.Institute of Image Processing & Pattern RecognitionShanghai Jiaotong UniversityShanghaiChina
  2. 2.Dept. of InformationSouthern Yangtse UniversityJiangsuChina

Personalised recommendations