OPTIMASI JUMLAH CLUSTER UNTUK ANALISIS PENJUALAN BARANG KOSMETIK MENGGUNAKAN K-MEDOIDSThis research aims to optimize the number of clusters in cosmetic product sales analysis using the K-Medoids algorithm on KTT store sales data. To determine the optimal number of clusters, we applied the elbow method, silhouette analysis, and gap statistic using Python during data analysis and algorithm implementation. The analyzed dataset includes monthly sales data with attributes such as number of units sold, unit price, and total revenue. In the three methods, each has different characteristics so as to produce different values of the optimum number of clusters for the dataset used, the difference in determining the optimal number of clusters does not mean aborting or weakening between one method and another, but can be used to evaluate one and the other. Where in this study the optimal number of clusters is determined by considering the results of each method by taking the middle value. There are 2 categories of product groupings that have the most sales and products that sell less. There are 3 product categories that have the most sales, namely wardah, somethinc, and focallure. While the product kategoris that sell less are 2 kategoris, namely instaperfect and dazzle me. Keywords - Clustering, K-Medoids, Elbow Method, Silhouette Analysis, Gap Statistic.

Rindiawan, Selvina Zalfa (2024) OPTIMASI JUMLAH CLUSTER UNTUK ANALISIS PENJUALAN BARANG KOSMETIK MENGGUNAKAN K-MEDOIDSThis research aims to optimize the number of clusters in cosmetic product sales analysis using the K-Medoids algorithm on KTT store sales data. To determine the optimal number of clusters, we applied the elbow method, silhouette analysis, and gap statistic using Python during data analysis and algorithm implementation. The analyzed dataset includes monthly sales data with attributes such as number of units sold, unit price, and total revenue. In the three methods, each has different characteristics so as to produce different values of the optimum number of clusters for the dataset used, the difference in determining the optimal number of clusters does not mean aborting or weakening between one method and another, but can be used to evaluate one and the other. Where in this study the optimal number of clusters is determined by considering the results of each method by taking the middle value. There are 2 categories of product groupings that have the most sales and products that sell less. There are 3 product categories that have the most sales, namely wardah, somethinc, and focallure. While the product kategoris that sell less are 2 kategoris, namely instaperfect and dazzle me. Keywords - Clustering, K-Medoids, Elbow Method, Silhouette Analysis, Gap Statistic. Other thesis, Universitas Siliwangi.

Text
1. COVER.pdf
Download (58kB)

[thumbnail of 2. LEMBAR PENGESAHAN TUGAS AKHIR.pdf]

Text
2. LEMBAR PENGESAHAN TUGAS AKHIR.pdf
Download (110kB)

[thumbnail of 3. PENGESAHAN PENGUJI SIDANG TUGAS AKHIR.pdf]

Text
3. PENGESAHAN PENGUJI SIDANG TUGAS AKHIR.pdf
Download (324kB)

[thumbnail of 4. LEMBAR PERNYATAAN TUGAS AKHIR.pdf]

Text
4. LEMBAR PERNYATAAN TUGAS AKHIR.pdf
Download (124kB)

Text
5. ABSTRACT.pdf
Download (29kB)

Text
6. ABSTRAK.pdf
Download (75kB)

[thumbnail of 7. MOTTO DAN PERSEMBAHAN.pdf]

Text
7. MOTTO DAN PERSEMBAHAN.pdf
Download (115kB)

Text
8. KATA PENGANTAR.pdf
Download (60kB)

Text
9. DAFTAR ISI.pdf
Download (59kB)

Text
10. DAFTAR GAMBAR.pdf
Download (33kB)

Text
11. BAB I.pdf
Download (111kB)

Text
12. BAB II.pdf
Download (220kB)

Text
13. BAB III.pdf
Download (59kB)

Text
14. BAB IV.pdf
Restricted to Repository staff only
Download (597kB)

Text
15. BAB V.pdf
Restricted to Repository staff only
Download (34kB)

Text
16. DAFTAR PUSTAKA.pdf
Download (171kB)

[thumbnail of TA_207006095_Selvina Zalfa Rindiawan.pdf]

Text
TA_207006095_Selvina Zalfa Rindiawan.pdf
Restricted to Repository staff only
Download (1MB)

Abstract

This research aims to optimize the number of clusters in cosmetic product sales analysis using the K-Medoids algorithm on KTT store sales data. To determine the optimal number of clusters, we applied the elbow method, silhouette analysis, and gap statistic using Python during data analysis and algorithm implementation. The analyzed dataset includes monthly sales data with attributes such as number of units sold, unit price, and total revenue. In the three methods, each has different characteristics so as to produce different values of the optimum number of clusters for the dataset used, the difference in determining the optimal number of clusters does not mean aborting or weakening between one method and another, but can be used to evaluate one and the other. Where in this study the optimal number of clusters is determined by considering the results of each method by taking the middle value. There are 2 categories of product groupings that have the most sales and products that sell less. There are 3 product categories that have the most sales, namely wardah, somethinc, and focallure. While the product kategoris that sell less are 2 kategoris, namely instaperfect and dazzle me.

Keywords - Clustering, K-Medoids, Elbow Method, Silhouette Analysis, Gap Statistic.

Item Type:	Thesis (Other)
Subjects:	T Technology > T Technology (General)
Divisions:	Fakultas Teknik > Teknik Informatika
Depositing User:	user1 user1 user1
Date Deposited:	14 Jan 2026 02:30
Last Modified:	14 Jan 2026 02:30
URI:	https://repositori.unsil.ac.id/id/eprint/4025

Actions (login required)

: View Item