Implementasi Sistem Rekomendasi Stok Produk dengan Algoritma K-Means Berdasarkan Data Penjualan (Studi Kasus: Toko Ajeng Ina)

Indhama, Zanna Chobita Majesty Ayu (2026) Implementasi Sistem Rekomendasi Stok Produk dengan Algoritma K-Means Berdasarkan Data Penjualan (Studi Kasus: Toko Ajeng Ina). Undergraduate thesis, Universitas Pembangunan Nasional "Veteran" Jawa Timur.

	Text (Cover) 19082010009.-cover.pdf Download (4MB)
	Text (Bab 1) 19082010009.-bab1.pdf Download (250kB)
	Text (Bab 2) 19082010009.-bab2.pdf Restricted to Repository staff only until 31 July 2028. Download (379kB)
	Text (Bab 3) 19082010009.-bab3.pdf Restricted to Repository staff only until 31 July 2028. Download (328kB)
	Text (Bab 4) 19082010009.-bab4.pdf Restricted to Repository staff only until 31 July 2028. Download (1MB)
	Text (Bab 5) 19082010009.-bab5.pdf Download (235kB)
	Text (Daftar Pustaka) 19082010009.-daftar pustaka.pdf Download (237kB)
	Text (Lampiran) 19082010009.-lampiran.pdf Restricted to Repository staff only until 31 July 2028. Download (323kB)

Abstract

Product inventory control is an essential part of store operations because it is closely related to sales continuity and cost efficiency. At Toko Ajeng Ina, stock recording and management are still performed manually. This practice often causes low-demand items to accumulate, while products with high sales levels may run out of stock. The situation emphasizes the importance of using historical sales transactions as the foundation for preparing inventory management recommendations. This research applies the K-Means algorithm to form product groups according to their sales characteristics, selects the most appropriate number of clusters through the Elbow Method, and develops a web-based product inventory recommendation system. The study follows the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology. The dataset was obtained from Toko Ajeng Ina's sales transactions from January to December 2025. After aggregation, it consisted of 350 product records with five attributes: Total Quantity, Sales Frequency, Total Sales, Total Restock, and Average Stock. The Elbow Method was applied to select the number of clusters, while clustering quality was assessed through the Silhouette Score and the Davies-Bouldin Index (DBI). The evaluation led to the selection of six clusters (K = 6) by considering the recommendations produced by the Elbow Method and the Davies-Bouldin Index, along with a reasonably adequate Silhouette Score. The model obtained a Silhouette Score of 0.280 and a Davies-Bouldin Index of 1.033. These values indicate that within-cluster compactness and separation between groups were at an acceptable level. The developed system automates the clustering procedure and presents the analytical results through an interactive dashboard, allowing the store owner to recognize product characteristics more easily and formulate effective inventory management strategies.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Diyasa, I Gede Susrama Mas	NIDN0019067008	igsusrama.if@upnjatim.ac.id
Thesis advisor	Putra, Agung Brastama	NIDN0024118503	agungbp.si@upnjatim.ac.id

Subjects:

Q Science > QA Mathematics > QA75 Electronic computers. Computer science

Divisions:

Faculty of Computer Science > Departemen of Information Systems

Depositing User:

Zanna Z Indhama

Date Deposited:

04 Aug 2026 03:19

Last Modified:

04 Aug 2026 03:41

URI:

https://repository.upnjatim.ac.id/id/eprint/28726