Behavior-Based Telecommunication Churn Prediction with Neural Network Approach

Zhang, Yongbin; Liang, Ronghua; Li, Yeli; Zheng, Yanying; Berry, Michael

doi:10.1109/isccs.2011.89

Cited by 20 publications

(14 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The first step in predictive modeling is the acquisition and preparation of data. Having the correct data is as important as having the correct method [2].…”

Section: Methodsmentioning

confidence: 99%

“…This is the major cause of the subscribers leaving one network and moving to another one which suits their needs. According to telecom market, the process of subscribers (either prepaid or post paid) switching from one service provider is called "customer churn" [2]. If churning continues to happen for any telecom industry, it would lead to the great loss of revenue to the company.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Applications of Data Mining Techniques in Telecom Churn Prediction

Umayaparvathi¹,

Iyakutti²

2012

IJCA

View full text Add to dashboard Cite

In this competitive world, business becomes highly saturated. Especially, the field of telecommunication faces complex challenges due to a number of vibrant competitive service providers. Therefore it has become very difficult for them to retain existing customers. Since the cost of acquiring new customers is much higher than the cost of retaining the existing customers, it is the time for the telecom industries to take necessary steps to retain the customers to stabilize their market value. This paper explores the application of data mining techniques in predicting likely churners and the impact of attribute selection on identifying the churn. It also compares the efficiency of Decision tree and Neural Network classifiers and lists their performances.

show abstract

“…The first step in predictive modeling is the acquisition and preparation of data. Having the correct data is as important as having the correct method [2].…”

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Applications of Data Mining Techniques in Telecom Churn Prediction

Umayaparvathi¹,

Iyakutti²

2012

IJCA

View full text Add to dashboard Cite

show abstract

“…Untuk mencapai kinerja yang baik dan akurasi yang tinggi jumlah fitur yang digunakan untuk memprediksi customer churn cukup enam sampai dengan delapan fitur. Tujuan pemilihan masukan membantu menghapus masukan yang tidak relevan, menghapus masukan yang bergantung dengan masukan lainnya, sehingga pembuatan model lebih ringkas, transparan dan mengurangi waktu untuk pembentukan model [6].…”

Section: Pendahuluanunclassified

Ig-KNN Untuk Prediksi Customer Churn Telekomunikasi

Arifin

2015

Simet

View full text Add to dashboard Cite

ABSTRAKIG-KNN merupakan gabungan dari algotitma pemilihan fitur information gain dengan algoritma klasifikasi KNN, kedua algoritma ini diharapkan dapat meningkatkan akurasi dalam memprediksi customer churn telekomunikasi. Prediksi customer churn telekomunikasi merupakan kebutuhan yang sangat penting bagi kelangsungan hidup perusahaan telekomunikasi, dimana dengan banyaknya pelanggang yang meninggalkan perusaan maka perusahaan berpeluang untuk merugi. Mendeteksi pelanggan yang berpeluang meninggalkan perusahaan sejak dini perusahaan akan mendapatkan keuntungan 10 kali, karena biaya untuk mempertahankan pelanggan lebih murah 10 kali lipat dibanding dengan mecari pelanggan baru. Berdasarkan hasil penelitian ini prediksi customer churn telekomunikasi dengan menggunakan IG-KNN menunjukkan akurasi yang lebih baik meski dengan nilai k yang berbedabeda bila dibandingkan dengan prediksi customer churn telekomunikasi dengan menggunkan KNN tanpa fitur seleksi Information Gain, adapun peningkatan akurasi dari k1 sampai dengan k11 sebesar 1,7%.Kata kunci: information gain, KNN, customer churn telekomunikasi. ABSTRACT IG-KNN is a combination of information gain feature selection algorithm with KNN classification algorithm, the second algorithm is expected to improve the accuracy of predicting customer churn in telecommunications. Telecommunication customer churn prediction is a very important requirement PENDAHULUANCustomer Churn didefinisikan sebagai kecenderungan pelanggan untuk berhenti melakukan bisnis dengan sebuah perusahaan [1]. Hal ini telah menjadi isu penting yang merupakan salah satu tantangan utama oleh banyak perusahaan di era global ini dan harus dihadapinya.[2] mengatakan untuk memperoleh pelanggan baru memerlukan biaya hingga 10 kali lipat lebih mahal dibandingkan biaya untuk mempertahankan pelanggan yang ada. Mahalnya untuk memperoleh pelanggan baru tentunya perusahaan akan lebih memilih mempertahankan pelanggan. Berdasarkan fakta tersebut maka banyak perusahaan sekarang lebih beralih untuk mempertahankan pelanggan dan menghindari churn pelanggan.Prediksi customer churn telekomunikasi menggunakan KNN yang dilakukan [3] menghasilkan nilai akurasi 88% pada nilai K 5 keatas, dengan menggunakan algoritma fitur seleksi diharapkan akan meningkatkan akurasi prediksi selain itu dapat diketahui fitur apa saja yang tidak dibutuhkan dalam memprediksi customer churn telekomunikasi.Pemilihan Subset fitur, bersama dengan pengaturan parameter dalam prosedur pelatihan SVM secara signifikan mempengaruhi akurasi klasifikasi [4]. Jumlah fitur yang banyak (high dimensional) akan memberikan masukan yang berbeda-beda. Pada prakteknya, atribut yang berlebihan tidak akan memberikan hasil yang signifikan pada proses pelatihan tetapi justru akan menyebabkan over-fit atau irrelevan dan redundan atribut yang akan menyulitkan algoritma proses pelatihan. Masih sedikit penelitian yang berfokus pada pemilihan fitur untuk mempediksi customer churn telekomunikasi [5]. Untuk mencapai kinerja yang baik dan akurasi yang tinggi jumlah fitur yang...

show abstract

“…Yongbin Zhang et al, [22] attempted to develop behavior-based telecommunication churn prediction system with artificial neural network approach (SOM) and used only customer service usage information for prediction. A.A.Khan et al, [23] used decision tree, logistic regression and neural network to predict churn and suggested that demographic features have the lowest affect on the churn prediction when compared to the billing and usage details.…”

Section: Customer Churnmentioning

confidence: 99%

“…Since the customer is the major source of profit, a method to promptly manage customer churn gains vital significance for the survival and development of any telecommunication company. For many telecoms companies, figuring out how to deal with Churn is turning out to be the key for continued existence of their organizations [20,22].…”

Section: Introductionmentioning

confidence: 99%

A Survey on Churn Prediction Techniques in Communication Sector

Kamalraj¹,

Malathi²

2013

IJCA

View full text Add to dashboard Cite

The speedy augmentation of the market in every sector is leading to superior subscriber base for service providers. Added competitors, novel and innovative business models and enhanced services are increasing the cost of customer acquisition. In such a tedious set up service providers have realized the importance of retaining the on hand customers. It is therefore mandatory for the service providers to inhibit churn-a phenomenon which states that customer wishes to quit the service of the company. To prevent the churn many approaches are used by the researchers. This paper reviews the different approaches used by researchers not only in communication sector but also other sectors which highly depends on customer participation.

show abstract

Behavior-Based Telecommunication Churn Prediction with Neural Network Approach

Cited by 20 publications

References 11 publications

Applications of Data Mining Techniques in Telecom Churn Prediction

Applications of Data Mining Techniques in Telecom Churn Prediction

Ig-KNN Untuk Prediksi Customer Churn Telekomunikasi

A Survey on Churn Prediction Techniques in Communication Sector

Contact Info

Product

Resources

About