Analisa Perbandingan Metode Prediksi Pelanggan Churn Dengan Penerapan Data Mining

Rozie, Ach. Nafila (2019) Analisa Perbandingan Metode Prediksi Pelanggan Churn Dengan Penerapan Data Mining. Undergraduate thesis, Institut Teknologi Sepuluh November.

[thumbnail of 02411540000113 - UNDERGRADUATE_THESES.pdf]
Preview
Text
02411540000113 - UNDERGRADUATE_THESES.pdf - Accepted Version

Download (4MB) | Preview

Abstract

Pelanggan merupakan salah satu unsur penting dalam memastikan suatu preses bisnis perusahaan agar dapat terus bertahan serta bersaing dengan kompetitor. Berdasarkan kondisi dilapangan, pelanggan dapat sepenuhnya loyal menggunakan layanan jasa/produk suatu perusahaan atau berpindah menggunakan layanan jasa/produk dari perusahaan lain (churn). Hasil penelitian juga menunjukkan bahwa dengan menurunkan tingkat pelanggan berhenti menggunakan layanan jasa/produk dari suatu perusahaan dapat meningkat pendapat perusahaan hingga 95%. PT X sebagai perusahaan yang bergerak dibidang telekomunikasi memandang bahwa mempertahankan pelanggan jauh lebih penting mengingat biaya yang dibutuhkan lebih murah jika dibandingkan dengan mengakuisisi pelanggan baru. Maka dari itu diperlukan suatu sistem prediksi pelanggan yang mampu mengidentifikasi apakah suatu pelanggan berhenti menggunakan layanan dari PT X atau tidak. Hal tersebut dapat dicapai salah satu caranya dengan pendekatan data mining. Teknik data mining yang dipilih untuk membangun model prediksi pada penelitian ini adalah teknik klasifikasi dengan Regresi Logistik, Naïve Bayes dan Support Vector Machine. Dari hasil komparasi model, diperoleh bahwa model terbaik menggunakan pendekatan Support Vector Machine dengan kernel RBF. Akurasi serta recall dari model tersebut dalam memprediksi pelanggan churn masing-masing mencapai 89.63% dan 89.79%. Uji coba sistem yang dilakukan menggunakan data pelanggan dari PT X di area kerja Bogor.
=================================================================================================
The customer is one of the important elements in ensuring a company can
survive and compete with competitors. Based on the field conditions, customers can
be fully loyal to use the services / products of a company or move to use services /
products from other companies (churn). The results of the study also show that by
reducing the level of customers stopping using services / products from a company
can increase the company's profit up to 95%. PT X as a telecommunications
company considers that retaining customers is far more important compared to
acquiring new customers. Therefore, they need a customer prediction system that is
able to identify whether a customer has stopped using services from PT X or not.
This can be achieved by using a data mining approach. The data mining techniques
chosen to build the prediction model in this study are classification techniques with
Logistic Regression, Naïve Bayes and Support Vector Machine. From the results of
the model comparison and parameters tuning, it was found that the best model used
the Support Vector Machine with RBF kernel approach. Accuracy and recall values
of the model reaching 89.63% and 89.79% respectively. System testing is carried
out using customer data from PT X in the Bogor work area.

Item Type: Thesis (Undergraduate)
Additional Information: RSI 006.312 Roz a-1 2019
Uncontrolled Keywords: Data Mining, Churn, Logistic Regression, Naïve Bayes Classifier, Support Vector Machine
Subjects: Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science)
T Technology > T Technology (General) > T174 Technological forecasting
T Technology > T Technology (General) > T57.5 Data Processing
T Technology > T Technology (General) > T58.62 Decision support systems
Divisions: Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26201-(S1) Undergraduate Thesis
Depositing User: Ach. Nafila Rozie
Date Deposited: 16 Sep 2021 20:02
Last Modified: 29 Apr 2024 11:14
URI: http://repository.its.ac.id/id/eprint/60298

Actions (login required)

View Item View Item