Please use this identifier to cite or link to this item:
https://repositori.uma.ac.id/handle/123456789/29227| Title: | Optimasi Penentuan Lokasi Tiang ODP Berdasarkan Clustering Pelanggan Menggunakan Algoritma K Means |
| Other Titles: | Optimization of ODP Pole Location Determination Based on Customer Clustering Using the K Means Algorithm |
| Authors: | Saputra, Mhd. Ramadhan |
| metadata.dc.contributor.advisor: | Muliono, Rizki |
| Keywords: | ISP;ODP;elbow;k-means clustering;DBI;silhouette score |
| Issue Date: | 17-Sep-2025 |
| Publisher: | Universitas Medan Area |
| Series/Report no.: | NPM;218160004 |
| Abstract: | Teknologi dan informasi yang terus berkembang mempengaruhi berbagai aspek kehidupan manusia, termasuk meningkatnya ketergantungan terhadap layanan internet. ISP (Internet Service Provider) menjadi penyedia layanan internet melalui berbagai metode distribusi jaringan, salah satunya adalah Wireless Fidelity (Wi-Fi). Untuk mendukung penyediaan Wi-Fi, diperlukan infrastruktur berupa ODP (Optical Distribution Point) sebagai pendistribusian jaringan ke masing-masing pelanggan. Agar memastikan keamanan ODP, diperlukan tiang sebagai tumpuannya. Namun, lokasi tiang ODP yang tidak strategis mengakibatkan teknisi kesulitan melakukan pengecekan sebelum pemasangan baru dan melakukan pemeliharaan terhadap jaringan tersebut. Teknisi diharuskan untuk mencari tiang ODP di daerah alamat pelanggan agar memproses pemasangan baru serta pemeliharaan. Selain itu, teknisi juga tidak dapat memastikan lokasi tiang ODP tanpa mengecek ODP yang berada disisi tiang. Berdasarkan permasalahan tersebut, diusulkan penelitian ini yang mengelompokkan data pelanggan melalui metode k-means. Kemudian untuk mengoptimalkan proses clustering, diperlukan metode elbow dalam menentukan jumlah cluster. Selanjutnya DBI (Davies-Bouldin Index) dan silhouette score digunakan untuk mengukur kualitas hasil cluster. Hasil yang diperoleh, sistem mampu melakukan clustering dan penempatan tiang ODP dengan berbagai variasi jumlah data. Penentuan jumlah cluster dengan elbow menghasilkan cluster optimal berjumlah 5 dengan nilai DBI sebesar 0,55 dan silhouette score sebesar 0,42 pada 500 data yang dikategorikan cukup baik. Sementara itu, penentuan jumlah cluster menggunakan k-means (tanpa elbow) menghasilkan nilai DBI sebesar 0,78 dan silhouette score sebesar 0,34 pada 500 data. Hasilnya, metode elbow mampu melengkapi metode k-means dalam menghasilkan jumlah cluster optimal. Sehingga penelitian ini mampu memberikan kontribusi dalam penentuan lokasi tiang ODP yang diharapkan meminimalkan penggunaan sumber daya. Technology and information that continues to develop affects various aspects of human life, including increasing dependence on internet services. ISP (Internet Service Provider) provides internet services through various network distribution methods, one of which is Wireless Fidelity (Wi-Fi). To support the provision of Wi-Fi, infrastructure is needed in the form of an ODP (Optical Distribution Point) to distribute the network to each customer. In order to ensure the safety of the ODP, a pole is needed as a support. However, the location of the ODP poles is not strategic, making it difficult for technicians to carry out checks before new installations and carrying out maintenance on the network. Technicians are required to look for ODP poles in the customer's address area to process new installations and maintenance. Apart from that, technicians cannot confirm the location of the ODP pole without checking the ODP on the side of the pole. Based on these problems, this research is proposed which groups customer data using the k-means method. Then, to optimize the clustering process, the elbow method is needed to determine the number of clusters. Furthermore, DBI (Davies-Bouldin Index) and silhouette score are used to measure the quality of cluster results. The results obtained show that the system is able to perform clustering and placement of ODP poles with various variations in the amount of data. Determining the number of clusters using elbows produces an optimal cluster of 5 with a DBI value of 0.55 and a silhouette score of 0.42 on 500 data which is categorized as quite good. Meanwhile, determining the number of clusters using k-means (without elbows) produces a DBI value of 0.78 and a silhouette score of 0.34 on 500 data. As a result, the elbow method is able to complement the k-means method in producing the optimal number of clusters. So this research is able to contribute to determining the location of ODP poles which is expected to minimize resource use. |
| Description: | 81 Halaman |
| URI: | https://repositori.uma.ac.id/handle/123456789/29227 |
| Appears in Collections: | SP - Informatic Engineering |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 218160004 - Mhd. Ramadhan Saputra - Chapter IV.pdf Restricted Access | Chapter IV | 2.43 MB | Adobe PDF | View/Open Request a copy |
| 218160004 - Mhd. Ramadhan Saputra - Fulltext.pdf | Cover, Abstract, Chapter I, II, III, V, Bibliography | 1.78 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.