Favian, Iqbal Grady (2020) Analisis Pelanggan Pada Industri Alat Kesehatan Berbasis Clustering Dan Classification. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
09211750055003-master_thesis.pdf Download (3MB) | Preview |
Abstract
Memelihara hubungan dengan pelanggan dan melihat pola pelanggan dalam pasar merupakan sebuah cara perusahaan untuk tetap kompetitif disaat ini. Cara ini dikenal dengan Customer relationship management (CRM). Untuk menghadapi semua itu, perusahaan dihadapkan dengan beberapa segmen pelanggan yang memiliki berbagai macam perbedaan antara lain kebiasaan dalam membeli barang. Jadi setiap segmen diperlukan cara-cara yang berbeda dalam membangun hubungan dengan pelanggan. Untuk mencapai hal ini maka segmentasi pelanggan merupakan langkah yang harus ditempuh terutama di industry kesehatan saat ini. Untuk mendapat segmentasi yang sesuai, maka dalam penelitian ini digunakan LRFM (Length, Recency, Frequency, Monetary) sebagai atribut dasar dalam segmentasi. Penelitian ini menggunakan K-Means sebagai metode yang digunakan untuk proses segmentasi pelanggan berdasarkan LRFM. Kemudian penelitian ini juga menggunakan Decision Tree sebagai sistem pendukung keputusan dan penerapan IF-THEN rules. Pada penelitian ini dilakukan proses clustering yang terbagi menjadi 10 proses uji coba. Kemudian dari 10 proses uji tersebut dilihat berdasarkan validasi Davies-bouldin index didapat nilai.K=4 merupakan nilai yang optimal. Segmenyang terbentuk dibagi menjadi 4 segmen sesuai hasil clustering. Segmen tersebut dilakukan pengukuran akurasi menggunakan decision-tree untuk melihat keakurasian setiap segmen. Setelah itu akan dibentuk IF-Then rules untuk menetapkan rule berdasarkan segmen yang telah terbentuk pada proses clustering. Kemudian dilakukan analisis strategi sesuai kondisi segmen yang telah terbentuk berdasarkan decision-rule yang telah terbentuk.
===============================================================================================================================
Maintaining relationships with customers and seeing the pattern of customers in the market is a way for companies to remain competitive at this time. This method is known as Customer relationship management (CRM). To deal with all that, the company is faced with several customer segments that have a variety of differences, including habits in buying goods. So each segment needs different ways to build relationships with customers. To achieve this, customer segmentation is a step that must be taken, especially in the current health industry. To get the appropriate segmentation, in this study used LRFM (Length, Recency, Frequency, Monetary) as a basic attribute in segmentation. This study uses K-Means as the method used for the customer segmentation process based on LRFM. Then this study also uses the Decision Tree as a decision support system and the application of IF-THEN rules. In this research, the clustering process is divided into 10 trial processes. Then from the 10 trial processes seen value of Davies-bouldin index validation, the value of K = 4 is the optimal value. Segments are formed into 4 segments according to the results of the grouping. The segment is measured for approval using a decision tree to see the accuracy of each segment. Then will be form into IF-Then rules to set rules based on segments that have been formed in the clustering process. Then the strategy analysis is carried out in accordance with the segments that have been formed based on the decisions that have been formed.
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTMT 519.53 Fav a-1 |
| Uncontrolled Keywords: | Clustering, Classification, Decision Tree, IF-Then rules, Algortima KMeans |
| Subjects: | T Technology > T Technology (General) |
| Divisions: | Faculty of Creative Design and Digital Business (CREABIZ) > Technology Management > 61101-(S2) Master Thesis |
| Depositing User: | Iqbal Grady Favian |
| Date Deposited: | 17 Mar 2025 03:46 |
| Last Modified: | 17 Mar 2025 03:46 |
| URI: | http://repository.its.ac.id/id/eprint/74820 |
Actions (login required)
 |
View Item |