Improving Fuzzy Kohonen Clustering Network  and Applying it on Iris Data and Medical Image Segmentation

Mohammad  Assaad

Article Details

Improving Fuzzy Kohonen Clustering Network and Applying it on Iris Data and Medical Image Segmentation

Authors:

Mohammad Assaad

Publisher:

Latakia University, Syria

Journal:

Latakia University Journal -Basic Sciences Series Vol.39 Issue 4, pp.0-0

Document Type:

Article

Publication Date:

1/2017

Document Language:

English

Document ID (Org No.):

2001524841

pISSN:

2663-4252

Abstract:

 In this research fuzzy kohonen clustering network has been tested after proposing a new relationship for the fuzzy parameter m in a clustering process. By studying the work of this algorithm, it shows that the most influential factor in fuzzy kohonen network behavior was fuzzy parameter m, This is because the update fuzzy parameterin each iteration automatically and effectively controls to distribution of learning rate and update the network weights to all data nodes at each. In this research the improved Fuzzy Kohonen Clustering Network (IFKCN) was tested on (Fisher’s IRIS) data, which is the standard data are testing the effectiveness of new algorithms, which represent data for three different classes of flowers. The proposed algorithm has  been tested in the field of medical image segmentation (magnetic resonance images of the brain) because of this application has great importance in the medical field. The proposed algorithm results were compared in the two areas of the former applications with two clustering famous K_means,(Fuzzy C-Means, FCM) and Fuzzy Kohonen Clustering Network (FKCN). The IFKCN algorithm results showed a good result comparing with the previous three algorithms in terms of the number of iterations and centers of clusters, mean square error ( MSE).

Article Details

Improving Fuzzy Kohonen Clustering Network and Applying it on Iris Data and Medical Image Segmentation

Authors: Mohammad Assaad

Publisher: Latakia University, Syria

Journal: Latakia University Journal -Basic Sciences Series Vol.39 Issue 4, pp.0-0

Document Type: Article

Publication Date: 1/2017

Document Language: English

Document ID (Org No.): 2001524841

pISSN: 2663-4252

Issues

Authors:

Mohammad Assaad

Publisher:

Latakia University, Syria

Journal:

Latakia University Journal -Basic Sciences Series Vol.39 Issue 4, pp.0-0

Document Type:

Article

Publication Date:

1/2017

Document Language:

English

Document ID (Org No.):

2001524841

pISSN:

2663-4252