Pengembangan Model Prediksi Harga Pupuk Bulanan: Studi Kasus Pengabdian Masyarakat Lab Algoritma Pemrograman Teknik Informatika ITS

Rachmadi, Hilmi Zharfan (2024) Pengembangan Model Prediksi Harga Pupuk Bulanan: Studi Kasus Pengabdian Masyarakat Lab Algoritma Pemrograman Teknik Informatika ITS. Project Report. [s.n.], [s.l.]. (Unpublished)

Text 5025201268-Project_Report.pdf - Accepted Version Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

Dalam dekade terakhir, industri pupuk global mengalami volatilitas harga yang signifikan yang diperparah oleh pandemi COVID-19 dan konflik geopolitik, seperti konflik Rusia-Ukraina. Di tengah volatilitas ini, PT X, salah satu BUMN produsen pupuk terbesar di Asia Tenggara, menghadapi tantangan dalam menentukan harga pupuk-pupuk mereka. Analisis prediktif terhadap harga pupuk menjadi penting untuk mengantisipasi dinamika pasar dan mempersiapkan strategi penjualan yang efektif. Oleh karena itu, dilakukan Kerja Praktik di Lab Algoritma dan Pemrograman dengan tujuan mengembangkan model prediksi harga pupuk amonia dan urea bulanan untuk PT X. Model prediksi dikembangkan dengan dua metode, yakni metode supervised machine learning, dan metode statistik. Mean Absolute Percentage Error (MAPE) digunakan sebagai metrik evaluasi utama dengan target MAPE di bawah 5% dari PT X. Hasil pengembangan model menunjukkan bahwa SARIMA terbaik untuk prediksi harga amonia, sedangkan SVR terbaik untuk urea. Prediksi jangka pendek berhasil mencapai target MAPE di bawah 5% untuk kedua jenis pupuk. Akan tetapi, prediksi jangka panjang hanya berhasil mencapai target untuk urea. Kesulitan prediksi jangka panjang dalam mencapai target MAPE disebabkan oleh berbagai faktor eksternal yang mempengaruhi fluktuasi harga pupuk. Berdasarkan hasil ini, kami merekomendasikan integrasi faktor eksternal sebagai variabel eksogen dalam model.
============================================================================================================================
Over the past decade, the global fertilizer industry has seen considerable price fluctuation, worsened by factors like the COVID-19 pandemic and geopolitical conflicts, such as the Russia-Ukraine dispute. PT X, a leading state-owned fertilizer manufacturer in Southeast Asia, is feeling the impact of these challenges in their fertilizer pricing. Predictive analysis of fertilizer prices has become vital to anticipate market changes and devise effective sales strategies. Consequently, work was conducted in the Algorithm and Programming Lab with the goal of creating a monthly price prediction model for ammonia and urea fertilizers for PT X. The prediction model was developed using two methods: supervised machine learning and the statistical method. The main evaluation metric was the Mean Absolute Percentage Error (MAPE), with a target of less than 5% for PT X. The results showed that SARIMA was the best for predicting ammonia prices, while SVR was best for urea. Short-term predictions achieved the target MAPE of less than 5% for both fertilizers. However, only urea met the target for long-term predictions. The difficulty in achieving the MAPE target for long-term predictions can be attributed to various external factors affecting fertilizer prices. Therefore, we recommend incorporating these external factors as exogenous variables in the model.

Item Type:	Monograph (Project Report)
Uncontrolled Keywords:	Harga Pupuk, Prediksi, Deret Waktu, SVR, SARIMA
Subjects:	S Agriculture > S Agriculture (General) > S633.5 Fertilizers T Technology > T Technology (General) > T174 Technological forecasting T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing
Divisions:	Faculty of Industrial Technology > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Hilmi Zharfan Rachmadi
Date Deposited:	08 Jan 2024 00:19
Last Modified:	08 Jan 2024 00:19
URI:	http://repository.its.ac.id/id/eprint/105387

Actions (login required)

View Item