PENGGEROMBOLAN PROVINSI DI INDONESIA BERDASARKAN HARGA DAGING AYAM RAS MENGGUNAKAN METODE K-MEDOIDS DENGAN JARAK DYNAMIC TIME WARPING

ANGELICA BILFRIDA RAJAGUKGUK, . (2026) PENGGEROMBOLAN PROVINSI DI INDONESIA BERDASARKAN HARGA DAGING AYAM RAS MENGGUNAKAN METODE K-MEDOIDS DENGAN JARAK DYNAMIC TIME WARPING. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text COVER.pdf Download (1MB)
	Text BAB I.pdf Download (89kB)
	Text BAB II.pdf Restricted to Registered users only Download (456kB) | Request a copy
	Text BAB III.pdf Restricted to Registered users only Download (159kB) | Request a copy
	Text BAB IV.pdf Restricted to Registered users only Download (941kB) | Request a copy
	Text BAB V.pdf Restricted to Registered users only Download (129kB) | Request a copy
	Text Daftar Pustaka.pdf Download (192kB)
	Text Lampiran.pdf Restricted to Registered users only Download (655kB) | Request a copy

Abstract

Harga daging ayam ras merupakan salah satu komoditas pangan strategis yang berpengaruh terhadap inflasi dan pemenuhan kebutuhan protein masyarakat di Indonesia. Perbedaan kondisi geografis, distribusi, serta faktor ekonomi menyebabkan pola pergerakan harga daging ayam ras antar provinsi tidak selalu sama sehingga diperlukan analisis penggerombolan untuk mengidentifikasi provinsi yang memiliki karakteristik pola harga yang serupa. Penelitian ini bertujuan untuk mengelompokkan provinsi di Indonesia berdasarkan data deret waktu harga bulanan daging ayam ras periode Januari 2019 sampai Desember 2025 menggunakan metode K-Medoids dengan ukuran jarak Dynamic Time Warping (DTW). Penelitian diawali dengan standarisasi data menggunakan z-score untuk menghilangkan pengaruh skala data. Selanjutnya dihitung matriks jarak menggunakan DTW yang mampu mengukur kemiripan pola deret waktu meskipun terjadi pergeseran waktu. Matriks jarak DTW kemudian digunakan sebagai masukan pada algoritma K-Medoids, sedangkan jumlah gerombol optimal ditentukan berdasarkan nilai Silhouette Coefficient tertinggi. Hasil penelitian menunjukkan bahwa jumlah gerombol optimal yang terbentuk adalah 2 gerombol. Gerombol pertama terdiri atas 19 provinsi, sedangkan gerombol kedua terdiri atas 11 provinsi. Kedua gerombol memiliki karakteristik pola fluktuasi harga yang berbeda, di mana provinsi dalam gerombol yang sama menunjukkan pola perubahan harga yang relatif serupa selama periode pengamatan. Nilai rata-rata Silhouette Coefficient yang diperoleh sebesar 0,3834, yang menunjukkan bahwa struktur penggerombolan yang dihasilkan termasuk dalam kategori weak structure. Dengan demikian, metode K-Medoids menggunakan ukuran jarak Dynamic Time Warping mampu mengelompokkan provinsi berdasarkan kemiripan pola pergerakan harga daging ayam ras selama periode penelitian. ***** Broiler chicken meat is one of the strategic food commodities that plays an important role in inflation and in meeting the protein needs of the Indonesian population. Differences in geographical conditions, distribution systems, and economic factors lead to varying patterns of broiler chicken meat prices across provinces, making cluster analysis necessary to identify provinces with similar price movement patterns. This study aims to cluster Indonesian provinces based on monthly time series data of broiler chicken meat prices from January 2019 to December 2025 using the K-Medoids method with the Dynamic Time Warping (DTW) distance measure. The analysis began with data standardization using the z-score transformation to eliminate the effect of different data scales. Subsequently, a DTW distance matrix was computed to measure the similarity of time series patterns despite possible temporal shifts. The resulting DTW distance matrix was then used as the input for the K-Medoids algorithm, while the optimal number of clusters was determined based on the highest Silhouette Coefficient value. The results showed that the optimal clustering solution consisted of two clusters. The first cluster contained 19 provinces, while the second cluster consisted of 11 provinces. The two clusters exhibited different price fluctuation characteristics, with provinces within the same cluster sharing relatively similar price movement patterns throughout the observation period. The average Silhouette Coefficient obtained was 0.3834, indicating that the resulting clustering structure falls into the weak structure category. Therefore, the K-Medoids method using the Dynamic Time Warping distance measure was able to cluster provinces based on the similarity of broiler chicken meat price movement patterns during the study period.

Item Type:	Thesis (Sarjana)
Additional Information:	1) Prof. Dr. Ir. Bagus Sumargo, M.Si. ; 2) Dania Siregar, S.Stat., M.SI.
Subjects:	Sains > Sains, Ilmu Pengetahuan Alam Sains > Matematika Sains > Statistika
Divisions:	FMIPA > S1 Statistika
Depositing User:	Angelica Bilfrida Rajagukguk .
Date Deposited:	19 Aug 2026 07:19
Last Modified:	19 Aug 2026 07:19
URI:	http://repository.unj.ac.id/id/eprint/71836

Actions (login required)

View Item