Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/166746| Title: | Model dan Analisis Prediksi Permintaan Ayam Potong di PT XYZ |
| Other Titles: | |
| Authors: | Machfud Marimin Salsabhila, An'nisa |
| Issue Date: | 2025 |
| Publisher: | IPB University |
| Abstract: | PT XYZ merupakan perusahaan distribusi yang tengah mengembangkan lini bisnis baru di bidang distribusi ayam potong beku. Dalam operasionalnya, perusahaan menghadapi tantangan berupa fluktuasi permintaan musiman dan keterbatasan pasokan dari pemasok tetap. Permintaan yang tinggi pada periode tertentu sering kali tidak diimbangi dengan kemampuan prediksi yang akurat, karena perusahaan masih menggunakan metode prediksi kualitatif dengan tingkat akurasi rendah. Penelitian ini bertujuan untuk mengevaluasi dan menentukan metode prediksi kuantitatif yang paling sesuai untuk digunakan di PT XYZ. Beberapa metode diuji menggunakan data historis dalam format bulanan, mingguan, dan harian. Hasil evaluasi menunjukkan bahwa metode Holt-Winter Multiplikatif menghasilkan nilai MAPE terkecil sebesar 22% pada data bulanan, namun metode Dekomposisi Multiplikatif dengan data mingguan dinilai lebih sesuai dengan sistem pengadaan ayam potong PT XYZ yang biasanya dilakukan secara mingguan, dengan nilai MAPE sebesar 48%. Metode ini mampu mereduksi tingkat kesalahan prediksi sebesar 15% dari yang sudah dilakukan perusahan dan mereduksi kerugian yang dialami perusahaan. PT XYZ is a distribution company currently expanding its business line into the frozen chicken distribution sector. In its operations, the company faces challenges such as seasonal demand fluctuations and limited supply from regular suppliers. High demand during certain periods is often not supported by accurate forecasting capabilities, as the company still relies on qualitative forecasting methods with low accuracy. This study aims to evaluate and determine the most suitable quantitative forecasting method for use at PT XYZ. Several methods were tested using historical data in monthly, weekly, and daily formats. The evaluation results show that the Holt-Winters Multiplicative method produces the lowest MAPE value of 22% for monthly data. However, the Multiplicative Decomposition method using weekly data is considered more appropriate for PT XYZ’s chicken procurement system, which is typically conducted on a weekly basis, with a MAPE value of 48%. This method is capable of reducing the forecasting error rate by 15% compared to the company’s current approach and helps to reduce the financial losses experienced by the company. |
| URI: | http://repository.ipb.ac.id/handle/123456789/166746 |
| Appears in Collections: | UT - Agroindustrial Technology |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| cover_F3401211021_d764a4236ec845b185343e32d65cace7.pdf | Cover | 1.86 MB | Adobe PDF | View/Open |
| fulltext_F3401211021_256450c7f63041e0840da52ca4dc3800.pdf Restricted Access | Fulltext | 3.31 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.