Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/171600Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Afendi, Farit Mochamad | - |
| dc.contributor.advisor | Susetyo, Budi | - |
| dc.contributor.author | Ain, Karimatu | - |
| dc.date.accessioned | 2025-11-27T23:40:09Z | - |
| dc.date.available | 2025-11-27T23:40:09Z | - |
| dc.date.issued | 2025 | - |
| dc.identifier.uri | http://repository.ipb.ac.id/handle/123456789/171600 | - |
| dc.description.abstract | Indonesia sebagai negara maritim memiliki potensi besar pada sektor perikanan tangkap. Volume produksinya bervariasi antar provinsi karena perbedaan sumber daya ikan dan kondisi geografis. Penelitian ini bertujuan mengelompokkan provinsi di Indonesia berdasarkan pola volume produksi perikanan tangkap menggunakan pendekatan penggerombolan deret waktu berbasis model. Setiap provinsi dimodelkan dengan SARIMA, kemudian dihitung jarak antar model menggunakan Jarak Piccolo. Proses penggerombolan dilakukan dengan metode hierarki, menghasilkan lima gerombol optimal dengan nilai Silhouette Score sebesar 0,626. Setiap gerombol direpresentasikan oleh satu prototipe yang dimodelkan ulang untuk melakukan peramalan. Hasil evaluasi menunjukkan bahwa pendekatan pemodelan berbasis gerombol mampu mengurangi jumlah model dari 30 menjadi 5, dengan nilai rata-rata MAPE peramalan sebesar 17%. Meskipun terdapat peningkatan error pada beberapa provinsi, sebagian besar masih memiliki selisih MAPE < 10% dibanding pemodelan individu. Temuan ini menunjukkan bahwa pendekatan gerombol tetap mampu memberikan akurasi peramalan yang layak dengan efisiensi komputasi yang lebih tinggi, serta dapat dimanfaatkan dalam penyusunan strategi pengelolaan perikanan tangkap lintas wilayah | - |
| dc.description.abstract | Indonesia, as a maritime country, has great potential in the capture fisheries sector. The production volume varies across provinces due to differences in fish resources and geographical conditions. This study aims to cluster provinces in Indonesia based on the time series pattern of capture fisheries production volume using a model-based approach. Each province was modeled using SARIMA, and Piccolo Distance was used to measure dissimilarity between models. Hierarchical clustering resulted in five optimal clusters with a Silhouette Score of 0.626. Each cluster was represented by a prototype, which was re-modeled for forecasting. Evaluation showed that the clustering-based approach reduced the number of models from 30 to 5, achieving an average MAPE 17%. Although certain provinces experienced higher forecasting errors, most showed a MAPE difference of less than 10% compared to individual modeling. These findings indicate that cluster-based forecasting can still provide acceptable accuracy with significantly reduced computational burden, making it a useful tool for regional fisheries management and planning. | - |
| dc.description.sponsorship | null | - |
| dc.language.iso | id | - |
| dc.publisher | IPB University | id |
| dc.title | Penggerombolan Volume Produksi Perikanan Tangkap di Indonesia Berbasis Model Deret Waktu | id |
| dc.title.alternative | Clustering of Capture Fisheries Production Volume in Indonesia Based on Time Series Model | - |
| dc.type | Skripsi | - |
| dc.subject.keyword | gerombol hierarki | id |
| dc.subject.keyword | jarak piccolo | id |
| dc.subject.keyword | peramalan | id |
| dc.subject.keyword | SARIMA | id |
| dc.subject.keyword | Perikanan tangkap | id |
| Appears in Collections: | UT - Statistics and Data Sciences | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| cover_G1401211001_ff5bc7e2c68b4d78b23cdbcfccb1262e.pdf | Cover | 514.19 kB | Adobe PDF | View/Open |
| fulltext_G1401211001_2603d9a71dd94d4699c401f1fa920f08.pdf Restricted Access | Fulltext | 1.4 MB | Adobe PDF | View/Open |
| lampiran_G1401211001_4486b33eacb14e06bc42b4e699209245.pdf Restricted Access | Lampiran | 1.02 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.