PENDEKATAN DATA MINING DALAM ANALISIS POLA PEMBELIAN DI PERUSAHAAN DISTRIBUSI MENGGUNAKAN AGGLOMERATIVE HIERARCHICAL CLUSTERING DAN FP-GROWTH

Khairunisa, Putri Anggellina (2025) PENDEKATAN DATA MINING DALAM ANALISIS POLA PEMBELIAN DI PERUSAHAAN DISTRIBUSI MENGGUNAKAN AGGLOMERATIVE HIERARCHICAL CLUSTERING DAN FP-GROWTH. S1 thesis, Universitas Jambi.

	Text ABSTRAK.pdf Download (69kB)
	Text BAB I.pdf Download (126kB)
	Text BAB V.pdf Download (116kB)
	Text COVER.pdf Download (66kB)
	Text DAFTAR PUSTAKA.pdf Download (111kB)
	Text FULL SKRIPSI.pdf Restricted to Repository staff only Download (3MB)
	Text HALAMAN PENGESAHAN.pdf Download (222kB)

Abstract

Penelitian ini menganalisis pola pembelian pelanggan di PT. Pinus Merah Abadi menggunakan metode Agglomerative Hierarchical Clustering (AHC) dan Frequent Pattern Growth (FP-Growth). Tujuan penelitian adalah mengelompokkan pelanggan berdasarkan riwayat transaksi dan mengidentifikasi pola produk yang sering dibeli bersama. Dataset terdiri dari 18.691 transaksi dari 3.336 pelanggan dan 174 produk selama periode Juli hingga September 2024. Proses diawali dengan pra-pemrosesan data, termasuk pembersihan dan normalisasi, dilanjutkan dengan pengelompokan menggunakan AHC dengan metrik Euclidean Distance dan metode Average Linkage. Hasil pengelompokan menunjukkan dua cluster optimal dengan skor siluet 0,7204. Pola pembelian ditemukan menggunakan FP-Growth dengan nilai minimum support 0,1 (10%) dan tingkat confidence 0,75 (75%). Cluster 0 yang terdiri dari 431 pelanggan menunjukkan pola pembelian terkuat: jika produk P300188 dan P300238 dibeli, maka produk P300098 juga dibeli. Sementara itu, Cluster 1 dengan 2.905 pelanggan menunjukkan pola pembelian terkuat yang sama: jika produk P300188 dan P300238 dibeli, maka produk P300098 juga dibeli. This study analyzes customer purchasing patterns at PT. Pinus Merah Abadi using the Agglomerative Hierarchical Clustering (AHC) and Frequent Pattern Growth (FP-Growth) methods. The aim of the research is to group customers based on transaction history and identify frequently co-purchased product patterns. The dataset consists of 18,691 transactions from 3,336 customers and 174 products during the period of July to September 2024. The process began with data preprocessing, including cleaning and normalization, followed by clustering using AHC with the Euclidean Distance metric and the Average Linkage method. The clustering results revealed two optimal clusters with a silhouette score of 0.7204. Purchasing patterns were discovered using FP-Growth, with a minimum support value of 0.1 (10%) and a confidence level of 0.75 (75%). Cluster 0, consisting of 431 customers, showed the strongest purchasing pattern: if products P300188 and P300238 are purchased, then product P300098 is also purchased. Meanwhile, Cluster 1, with 2,905 customers, demonstrated the same strongest purchasing pattern: if products P300188 and P300238 are purchased, then product P300098 is also purchased.

Type:	Thesis (S1)
Uncontrolled Keywords:	Agglomerative Hierarchical Clustering, FP-Growth, Data Mining, Customer Segmentation, and Purchase Pattern Analysis
Subjects:	L Education > L Education (General)
Divisions:	Fakultas Sains dan Teknologi > Sistem Informasi
Depositing User:	KHAIRUNISA
Date Deposited:	01 Jul 2025 07:59
Last Modified:	01 Jul 2025 07:59
URI:	https://repository.unja.ac.id/id/eprint/80865

Actions (login required)

View Item