Detecting SIM box fraud by using support vector machine and artificial neural network
Fraud in communication has been increasing dramatically due to the new modern technologies and the global superhighways of communication, resulting in loss of revenues and quality of service in telecommunication providers especially in Africa and Asia. One of the dominant types of fraud is SIM box...
Saved in:
| Main Authors: | , , , |
|---|---|
| 格式: | Article |
| 語言: | English |
| 出版: |
Penerbit UTM Press
2015
|
| 主題: | |
| 在線閱讀: | http://eprints.utm.my/id/eprint/58251/1/RoselinaSallehuddin2015_DetectingSIMBoxFraud.pdf http://eprints.utm.my/id/eprint/58251/ http://dx.doi.org/10.11113/jt.v74.2649 |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| id |
my.utm.58251 |
|---|---|
| record_format |
eprints |
| spelling |
my.utm.582512022-04-07T05:46:22Z http://eprints.utm.my/id/eprint/58251/ Detecting SIM box fraud by using support vector machine and artificial neural network Sallehuddin, Roselina Ibrahim, Subariah Mohd. Zain, Azlan Elmi, Abdikarim Hussein QA75 Electronic computers. Computer science Fraud in communication has been increasing dramatically due to the new modern technologies and the global superhighways of communication, resulting in loss of revenues and quality of service in telecommunication providers especially in Africa and Asia. One of the dominant types of fraud is SIM box bypass fraud whereby SIM cards are used to channel national and multinational calls away from mobile operators and deliver as local calls. Therefore it is important to find techniques that can detect this type of fraud efficiently. In this paper, two classification techniques, Artificial Neural Network (ANN) and Support Vector Machine (SVM) were developed to detect this type of fraud. The classification uses nine selected features of data extracted from Customer Database Record. The performance of ANN is compared with SVM to find which model gives the best performance. From the experiments, it is found that SVM model gives higher accuracy compared to ANN by giving the classification accuracy of 99.06% compared with ANN model, 98.71% accuracy. Besides, better accuracy performance, SVM also requires less computational time compared to ANN since it takes lesser amount of time in model building and training. Penerbit UTM Press 2015 Article PeerReviewed application/pdf en http://eprints.utm.my/id/eprint/58251/1/RoselinaSallehuddin2015_DetectingSIMBoxFraud.pdf Sallehuddin, Roselina and Ibrahim, Subariah and Mohd. Zain, Azlan and Elmi, Abdikarim Hussein (2015) Detecting SIM box fraud by using support vector machine and artificial neural network. Jurnal Teknologi, 74 (1). pp. 137-149. ISSN 0127-9696 http://dx.doi.org/10.11113/jt.v74.2649 DOI: 10.11113/jt.v74.2649 |
| institution |
Universiti Teknologi Malaysia |
| building |
UTM Library |
| collection |
Institutional Repository |
| continent |
Asia |
| country |
Malaysia |
| content_provider |
Universiti Teknologi Malaysia |
| content_source |
UTM Institutional Repository |
| url_provider |
http://eprints.utm.my/ |
| language |
English |
| topic |
QA75 Electronic computers. Computer science |
| spellingShingle |
QA75 Electronic computers. Computer science Sallehuddin, Roselina Ibrahim, Subariah Mohd. Zain, Azlan Elmi, Abdikarim Hussein Detecting SIM box fraud by using support vector machine and artificial neural network |
| description |
Fraud in communication has been increasing dramatically due to the new modern technologies and the global superhighways of communication, resulting in loss of revenues and quality of service in telecommunication providers especially in Africa and Asia. One of the dominant types of fraud is SIM box bypass fraud whereby SIM cards are used to channel national and multinational calls away from mobile operators and deliver as local calls. Therefore it is important to find techniques that can detect this type of fraud efficiently. In this paper, two classification techniques, Artificial Neural Network (ANN) and Support Vector Machine (SVM) were developed to detect this type of fraud. The classification uses nine selected features of data extracted from Customer Database Record. The performance of ANN is compared with SVM to find which model gives the best performance. From the experiments, it is found that SVM model gives higher accuracy compared to ANN by giving the classification accuracy of 99.06% compared with ANN model, 98.71% accuracy. Besides, better accuracy performance, SVM also requires less computational time compared to ANN since it takes lesser amount of time in model building and training. |
| format |
Article |
| author |
Sallehuddin, Roselina Ibrahim, Subariah Mohd. Zain, Azlan Elmi, Abdikarim Hussein |
| author_facet |
Sallehuddin, Roselina Ibrahim, Subariah Mohd. Zain, Azlan Elmi, Abdikarim Hussein |
| author_sort |
Sallehuddin, Roselina |
| title |
Detecting SIM box fraud by using support vector machine and artificial neural network |
| title_short |
Detecting SIM box fraud by using support vector machine and artificial neural network |
| title_full |
Detecting SIM box fraud by using support vector machine and artificial neural network |
| title_fullStr |
Detecting SIM box fraud by using support vector machine and artificial neural network |
| title_full_unstemmed |
Detecting SIM box fraud by using support vector machine and artificial neural network |
| title_sort |
detecting sim box fraud by using support vector machine and artificial neural network |
| publisher |
Penerbit UTM Press |
| publishDate |
2015 |
| url |
http://eprints.utm.my/id/eprint/58251/1/RoselinaSallehuddin2015_DetectingSIMBoxFraud.pdf http://eprints.utm.my/id/eprint/58251/ http://dx.doi.org/10.11113/jt.v74.2649 |
| _version_ |
1729703227720466432 |
| score |
13.251813 |