NURAHMAH, ASPIA (2026) ANALISIS SEGMENTASI TOKO MENGGUNAKAN ALGORITMA HDBSCAN PADA BUSINESS INTELLIGENCE DASHBOARD AROMETRIC CV MAGIX METRO MEDIA. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
COVER.pdf Download (813kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
BAB 1.pdf Restricted to Registered users only Download (171kB) |
|
|
Text (BAB II)
BAB 2.pdf Restricted to Registered users only Download (397kB) |
|
|
Text (BAB III)
BAB 3.pdf Restricted to Registered users only Download (263kB) |
|
|
Text (BAB IV)
BAB 4.pdf Restricted to Registered users only Download (455kB) |
|
|
Text (BAB V)
BAB 5.pdf Restricted to Registered users only Download (122kB) |
|
|
Text (DAFTAR PUSTAKA)
DAFTAR PUSTAKA.pdf Restricted to Registered users only Download (165kB) |
|
|
Text (LAMPIRAN)
LAMPIRAN.pdf Restricted to Registered users only Download (325kB) |
Abstract
The coffee industry in Indonesia has grown rapidly, including at CV Magix Metro Media, which oversees 43 coffee-selling MSME stores. The reliance on manual sales data management led to the development of the AROMETRIC Business Intelligence Dashboard, one of whose features is store segmentation using the HDBSCAN algorithm. This study aims to evaluate the quality of HDBSCAN clustering results using the Silhouette Score metric, analyze the characteristics of each store cluster, and identify stores classified as noise or outliers. The study applies a descriptiveanalytical quantitative approach to sales transaction data from 43 MSME stores over the 2021–2025 period, comprising a total sales volume of 1,504,300 units from 334,001 transactions. The research subjects were determined using total sampling, while the research object is the HDBSCAN segmentation output computed directly from the sales transaction data, covering six main features: sales volume, total transactions, product variety, average quantity per transaction, weekend ratio, and seasonal ratio. Product variety was found to be nearly constant across all stores but was nonetheless retained in the StandardScaler normalization, so the final segmentation is based on all six features. The HDBSCAN segmentation produced two main clusters, which were further segmented based on each store's Silhouette value into four categories: Cluster 0 (High-Priority Stores, 4 stores), Cluster 1 (Potential Stores, 31 stores), Cluster 2 (Stores Requiring Evaluation, 5 stores), and 3 stores identified as noise/outliers, each with distinct sales performance characteristics. These findings are expected to serve as a basis for CV Magix Metro Media's management in formulating more structured, datadriven business decisions. Keywords: Business Intelligence, HDBSCAN, store segmentation, clustering, Silhouette Score Industri kopi di Indonesia berkembang pesat, termasuk pada CV Magix Metro Media yang menaungi 43 toko UMKM penjual kopi. Pengelolaan data penjualan yang masih manual mendorong pengembangan sistem AROMETRIC Business Intelligence Dashboard, yang salah satu fiturnya adalah segmentasi toko menggunakan algoritma HDBSCAN. Penelitian ini bertujuan untuk mengevaluasi kualitas hasil clustering HDBSCAN menggunakan metrik Silhouette Score, menganalisis karakteristik setiap klaster toko, serta mengidentifikasi toko yang termasuk kategori noise atau outlier. Penelitian menggunakan pendekatan kuantitatif deskriptif analitik terhadap data transaksi penjualan 43 toko UMKM periode 2021–2025, dengan total volume penjualan 1.504.300 unit dari 334.001 transaksi. Subjek penelitian diambil menggunakan teknik total sampling, sedangkan objek penelitian adalah hasil segmentasi HDBSCAN yang dihitung langsung dari data transaksi penjualan, meliputi enam fitur utama: volume penjualan, total transaksi, variasi produk, rata-rata quantity per transaksi, weekend ratio, dan seasonal ratio. Variasi produk ditemukan bernilai nyaris konstan di seluruh toko namun tetap diikutsertakan dalam normalisasi StandardScaler, sehingga segmentasi akhir dibentuk berdasarkan keenam fitur tersebut. Hasil segmentasi HDBSCAN menghasilkan dua klaster utama yang kemudian disegmentasi lebih lanjut berdasarkan nilai Silhouette per toko menjadi empat kategori: Cluster 0 (Toko Prioritas Tinggi, 4 toko), Cluster 1 (Toko Potensial, 31 toko), Cluster 2 (Toko Perlu Evaluasi, 5 toko), dan 3 toko yang teridentifikasi sebagai noise/outlier, masing-masing dengan karakteristik performa penjualan yang berbeda. Hasil analisis ini diharapkan dapat menjadi dasar bagi manajemen CV Magix Metro Media dalam menyusun keputusan bisnis yang lebih terstruktur dan berbasis data. Kata Kunci: Business Intelligence, HDBSCAN, segmentasi toko, clustering, Silhouette Score
| Item Type: | Thesis (S1) |
|---|---|
| NIM/NIDN Creators: | 41822110018 |
| Uncontrolled Keywords: | Business Intelligence, HDBSCAN, segmentasi toko, clustering, Silhouette Score |
| Subjects: | 100 Philosophy and Psychology/Filsafat dan Psikologi > 150 Psychology/Psikologi > 153 Conscious Mental Process and Intelligence/Intelegensia, Kecerdasan Proses Intelektual dan Mental 300 Social Science/Ilmu-ilmu Sosial > 320 Political dan Government Science/Ilmu Politik dan Ilmu Pemerintahan > 322 Relation of The State of Organized Groups/Hubungan Negara dengan Kelompok Sosial yang Terorganisir > 322.3 Business and Industry/Bisnis dan Industri 500 Natural Science and Mathematics/Ilmu-ilmu Alam dan Matematika > 510 Mathematics/Matematika > 518 Numerical Analysis/Analisis Numerik, Analisa Numerik > 518.1 Algorithms/Algoritma |
| Divisions: | Fakultas Ilmu Komputer > Sistem Informasi |
| Depositing User: | khalimah |
| Date Deposited: | 29 Aug 2026 05:14 |
| Last Modified: | 29 Aug 2026 05:14 |
| URI: | http://repository.mercubuana.ac.id/id/eprint/103514 |
Actions (login required)
 |
View Item |
