CLUSTERING GSM INTERNET USAGE WITH SELF ORGANIZING MAPS

Öz

In the recent years, the mobile internet usage has been common in daily life by the development of GSM technologies. The internet provides a very easy and flexible connection everywhere, when the users need to use internet. It supports messaging, media, finance and lots of services throughout its connection. Data can be transferred by constant quota values of internet users. For all new quota requests, charging system makes queries. There is some kind of performance problems occurs when the applications of quota is adjusted due to low level users. If the quotas are planned to provide internet services regarding intensive level internet users, then the low level users become aggrieved. The aim of this study is to set up dynamic quota values according to users behaviors and thus ensure high performance income using most used artificial neural network in grouping studies as SOM (Self-Organizing Maps)

Anahtar Kelimeler

• GSM Charging, Chapman, SOM Neural Networks, Cluster Analysis

Kaynakça

1. Teuvo KOHONEN, Self-Organizing Maps ,Springer Series in Information Sciences, Hudaverdi BIRCAN,Metin ZONTUL,Ahmet Gurkan YUKSEK,A study of clustering exporting countries of Turkey using SOM Neural Networks, Ataturk University Journal of Economics and Administrative Sciences,Vol 20,No.2 ,2006, pp.219-238
2. Laurene FAUSETT,Fundamentals of Neural Networks, NJ:Prentice Hall,1994
3. A. Hamdy TAHA, Operation Research, NJ:Prentice Hall,1992
4. Howard ANTON, Chris RORRES, Elementary Linear Algebra Applications versions, John Wiley & Sons Inc Canada,1994
5. Nursel RUZGAR,Analysis of a company balance of payments with markov processes Dokuz Eylul University Journal of Graduate school of social sciences, Vol 5, No 1, 2003, pp164-179.