Claim Missing Document
Check
Articles

Found 2 Documents
Search

PERAMALAN PENJUALAN ASINAN BUAH MENGGUNAKAN HOLT WINTERS PYTHON DI TOKO BERBUAH DINGIN Mohammad Bagas Irianto; Widiyono Widiyono; Arief Soma Darmawan
Jurnal Pendidikan Teknologi Informasi (JUKANTI) Vol 8 No 2 (2025): JURNAL PENDIDIKAN TEKNOLOGI INFORMASI (JUKANTI) EDISI NOPEMBER 2025
Publisher : Universitas Citra Bangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37792/jukanti.v8i2.1882

Abstract

ABSTRAKPenelitian ini bertujuan untuk meramalkan penjualan asinan buah tahun 2025 menggunakan metode Holt–Winters Exponential Smoothing sebagai dasar perencanaan persediaan dan strategi bisnis. Data diperoleh dari catatan penjualan bulanan Januari 2023–Desember 2024 melalui metode dokumentasi. Tahap prapemrosesan dilakukan dengan smoothing untuk mereduksi fluktuasi acak, pembersihan outlier agar hasil lebih stabil, serta normalisasi untuk menyamakan skala data. Selanjutnya, data dibagi menjadi data latih untuk membangun model dan data uji untuk mengevaluasi akurasi. Holt–Winters dipilih karena mampu menangkap pola tren dan musiman dalam data penjualan. Evaluasi akurasi menggunakan Mean Absolute Error (MAE) dan Mean Absolute Percentage Error (MAPE). Hasil penelitian menunjukkan nilai MAE sebesar 20,18 dan MAPE 9,13%, yang termasuk kategori sangat akurat karena MAPE < 10%. Pola peramalan menggambarkan penurunan penjualan di pertengahan tahun serta peningkatan kembali di akhir tahun, sejalan dengan perilaku konsumsi masyarakat pada periode tertentu. Berdasarkan hasil penelitian, dapat disimpulkan bahwa metode Holt–Winters efektif digunakan untuk meramalkan penjualan asinan buah, serta bermanfaat bagi pelaku usaha dalam mengoptimalkan manajemen persediaan dan merumuskan strategi penjualan yang sesuai dengan pola musiman yang teridentifikasi. Kata kunci : forecasting, holt-winters, inventory, mae, mape, penjualan ABSTRACT This study aims to forecast fruit pickle sales in 2025 using the Holt–Winters Exponential Smoothing method as a basis for inventory planning and business strategy. The data were obtained from monthly sales records from January 2023 to December 2024 through documentation. Preprocessing included smoothing to reduce random fluctuations, outlier removal to ensure stability, and normalization to unify data scaling. The dataset was then divided into training data for model building and testing data for accuracy evaluation. The Holt–Winters method was chosen because it can capture both trend and seasonal components in sales data. Model accuracy was measured using Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE). The results show that the model achieved an MAE of 20.18 and a MAPE of 9.13%, which is categorized as highly accurate since the MAPE value is below 10%. The forecast pattern indicates a sales decline in the middle of the year followed by an increase towards the end of the year, reflecting seasonal consumer behavior. In conclusion, the Holt–Winters method is effective in forecasting fruit pickle sales and can support business owners in optimizing inventory management and designing sales strategies aligned with identified seasonal patterns. Keywords: forecasting, holt-winters, inventory, mae, mape, sales
Predicting Tablet Drug Expenditures Using Python-Based Facebook Prophet in Pharmaceutical Installations Kaka Rizki Maulana; Widiyono Widiyono; Arief Soma Darmawan
JURNAL TEKNOLOGI DAN OPEN SOURCE Vol. 8 No. 2 (2025): Jurnal Teknologi dan Open Source, December 2025
Publisher : Universitas Islam Kuantan Singingi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36378/jtos.v8i2.4907

Abstract

The increasing complexity of pharmaceutical logistics requires accurate forecasting to ensure drug availability and minimize the risk of stock shortages. This study aims to develop a forecasting model to predict monthly tablet drug expenditure in the Pharmacy Department. The research stages include problem identification, data collection from historical drug expenditure records, data pre-processing, and implementation of the forecasting model. The method used is Facebook Prophet, which was chosen for its ability to capture seasonal patterns, trends, and holidays in time series data. Model performance evaluation was conducted using Mean Absolute Percentage Error (MAPE) and Mean Absolute Error (MAE). The results showed that the model produced an MAE of 3,621.25 and a MAPE of 4.93%, indicating that the prediction accuracy level was in the good category. These findings prove that the Prophet method is capable of providing reliable results in drug expenditure forecasting. The results of this study are expected to support decision-making in drug requirement planning and improve the efficiency of pharmaceutical logistics management.