A Web Based Clustering Analysis Toolbox (WBCA) design Using MATLAB


SAVAŞ K., YILDIZ K.

2nd World Conference on Educational Sciences (WCES-2010), İstanbul, Türkiye, 4 - 08 Şubat 2010, cilt.2, ss.5276-5280 identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası: 2
  • Doi Numarası: 10.1016/j.sbspro.2010.03.860
  • Basıldığı Şehir: İstanbul
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.5276-5280
  • Marmara Üniversitesi Adresli: Evet

Özet

With the development technology, datas which are saved in databases are rising, useful methods and kind of algorithms for making employable data and are build up for processing. For the solution of this problem data mining tools come into existence, to which clustering algorithms belong. For the clustering area this paper supposed to aid education, the tool was designed for clustering algorithms used easy and effective, cut down of the time that is making programme by the students so with the using of data sets that belongs to students, results of the clustering that was supported graphical items shown with the help of user interface and clustering must be so effective and easy. Purpose of this paper, MATLAB software that is used common by the academic area, making web based toolbox using clustering algorithms for the user. This toolbox use kmeans, agnes, fuzzy cmeans algorithms that are exist in MATLAB software and available web interface for the evaluate clustering results with cluster validity criteria. By courtesy of the toolbox, web user that do not need to have MATLAB software and programming knowledge but only a web browser and they can load their own data into the web server then see the result by the chose of algorithm and download the results their own local computers. In this design MATLAB Web Server (MWS) was used that belong to MATLAB software. With the help of this tool web users can send the parameters to server via the internet and the computer that has on the MWS software, used this then working MWS that is developed in this paper, results can have seen with the web server (C) 2010 Elsevier Ltd. All rights reserved.