Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


case1058445974.pdf (2.25 MB)

ETD Abstract Container

Abstract Header

Competitive recurrent neural network model for clustering of multispectral data

Amartur, Sundar C.
http://rave.ohiolink.edu/etdc/view?acc_num=case1058445974

Abstract Details

1995, Doctor of Philosophy, Case Western Reserve University, Electrical Engineering.
Clustering and categorization constitute two important elements in the field of numerical taxonomy of data and find application in many engineering problems. In this thesis we have introduced a constraint satisfying neuron model for clustering of multi-spectral data. The proposed method overcomes some of the problems encountered with the Hopfield network. We have also compared this method with another neuron model known as the normalized mean field annealing or Potts neuron which also has the implicit property of satisfying the problem constraints. A detailed convergence analysis of the proposed model is given. The method was tested with a benchmark data, three sets of multi-dimensional magnetic resonance images, and a set of remotely sensed satellite images. An analysis of the performance of this method with maps representing the classification results are also presented. From the results we conclude that the proposed method is fast and provides accurate clustering of multi-dimensional data.
Yoshiyasu Takefugi (Advisor)
87 p.
Neural network model; Multispectral data

Recommended Citations

Citations

  • Amartur, S. C. (1995). Competitive recurrent neural network model for clustering of multispectral data [Doctoral dissertation, Case Western Reserve University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=case1058445974

    APA Style (7th edition)

  • Amartur, Sundar. Competitive recurrent neural network model for clustering of multispectral data. 1995. Case Western Reserve University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=case1058445974.

    MLA Style (8th edition)

  • Amartur, Sundar. "Competitive recurrent neural network model for clustering of multispectral data." Doctoral dissertation, Case Western Reserve University, 1995. http://rave.ohiolink.edu/etdc/view?acc_num=case1058445974

    Chicago Manual of Style (17th edition)

case1058445974
523
© 1995, all rights reserved.
This open access ETD is published by Case Western Reserve University School of Graduate Studies and OhioLINK.