IPB University Logo

SCIENTIFIC REPOSITORY

IPB University Scientific Repository collects, disseminates, and provides persistent and reliable access to the research and scholarship of faculty, staff, and students at IPB University

AI Repository
 
Building and Categories


      View Item 
      •   IPB Repository
      • Final Assignments
      • Undergraduate Final Assignments
      • UF - School of Data Science, Mathematic and Informatics
      • UF - Statistics and Data Sciences
      • View Item
      •   IPB Repository
      • Final Assignments
      • Undergraduate Final Assignments
      • UF - School of Data Science, Mathematic and Informatics
      • UF - Statistics and Data Sciences
      • View Item
      JavaScript is disabled for your browser. Some features of this site may not work without it.

      Penggerombolan dan Peramalan Harga Cabai Merah Provinsi di Indonesia dengan Pendekatan Ensemble Feature-based Clustering Time Series dan SARIMAX

      Thumbnail
      View/Open
      Cover (476.1Kb)
      Fulltext (1.335Mb)
      Lampiran (1.108Mb)
      Date
      2026
      Jenis/Type
      Skripsi
      Subtype
      Undergraduate Theses
      Author
      Hariyol
      Angraini, Yenni
      Soleh, Agus Mohamad
      Metadata
      Show full item record
      Abstract
      Cabai merah merupakan komoditas hortikultura strategis di Indonesia yang memiliki pergerakan harga fluktuatif dan volatilitas tinggi. Penelitian ini bertujuan menggerombolkan provinsi-provinsi di Indonesia berdasarkan karakteristik pergerakan harga cabai merah menggunakan pendekatan feature-based clustering time series, memodelkan data deret waktu harga cabai merah pada setiap gerombol yang terbentuk menggunakan pendekatan SARIMAX, serta meramalkan harga cabai merah pada masing-masing gerombol berdasarkan model terbaik yang diperoleh. Data yang digunakan berupa harga cabai merah mingguan tingkat provinsi yang bersumber dari Pusat Informasi Harga Pangan Strategis Nasional (PIHPS) Bank Indonesia pada periode Januari 2019 hingga Desember 2025. Prosedur penggerombolan dilakukan melalui ekstraksi fitur menggunakan Autocorrelation Function (ACF), penggerombolan hierarki dengan lima metode pautan, integrasi ensemble menggunakan Cluster-based Similarity Partitioning Algorithm (CSPA), dan algoritma Walktrap. Selanjutnya, pemodelan SARIMAX dilakukan dengan melibatkan hari besar keagamaan sebagai peubah eksogen, sedangkan evaluasi performa model menggunakan Mean Absolute Percentage Error (MAPE). Hasil penelitian menunjukkan terbentuknya empat gerombol konsensus optimal. Peubah eksogen hari besar keagamaan berpengaruh signifikan pada model prototype Gerombol 1 dan Gerombol 4, tetapi tidak signifikan pada Gerombol 2 dan Gerombol 3. Nilai MAPE data uji berkisar antara 9,53% hingga 16,61% dengan Gerombol 4 berada pada kategori sangat baik (MAPE < 10%) dan Gerombol 1, 2, serta 3 berada pada kategori baik (10–20%). Selain itu, MAPE pada tingkat prototype secara konsisten lebih kecil dibandingkan MAPE pada tingkat anggota gerombol. Integrasi pendekatan feature-based clustering time series dan SARIMAX terbukti efektif dalam menangkap dinamika harga antargerombol serta meningkatkan efisiensi komputasi karena hanya memerlukan empat model peramalan
       
      Red chili is a strategic horticultural commodity in Indonesia, characterized by highly fluctuating price movements and elevated volatility. This study aims to cluster Indonesian provinces based on red chili price movement characteristics using a feature-based clustering time series approach, model the time series data of red chili prices within each formed cluster using the SARIMAX approach, and forecast red chili prices for each cluster based on the best model obtained. The data used comprise weekly provincial level red chili prices from the National Strategic Food Price Information Center (PIHPS) of Bank Indonesia covering the period January 2019 to December 2025, with the clustering procedure involving feature extraction using the Autocorrelation Function (ACF), hierarchical clustering with five linkage methods, ensemble integration via the Cluster-based Similarity Partitioning Algorithm (CSPA) and the Walktrap algorithm, followed by SARIMAX modeling with major religious holidays as exogenous variables, and performance evaluation using the Mean Absolute Percentage Error (MAPE). The results indicate the formation of four optimal consensus clusters, with major religious holiday exogenous variables having a significant effect on the models for prototypes of Clusters 1 and 4 but not for Clusters 2 and 3, with MAPE values ranging from 9.53% to 16.61%, where Cluster 4 falls into the highly accurate category (MAPE < 10%) and Clusters 1, 2, and 3 fall into the good category (10%–20%), with prototype-level MAPE consistently lower than member level MAPE. The integrated feature-based clustering time series and SARIMAX approach proved effective in capturing inter provincial price dynamics and improving computational efficiency by requiring only four forecasting models
       
      URI
      http://repository.ipb.ac.id/handle/123456789/174617
      Collections
      • UF - Statistics and Data Sciences [166]

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository
        

       

      Browse

      All of IPB RepositoryCollectionsBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

      My Account

      Login

      Application

      google store

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository