Prediksi Penjualan Skincare Bulanan Menggunakan Arima, Sarima, Dan Prophet
DOI:
https://doi.org/10.31849/6tkw5817Keywords:
Time Series Forecasting, ARIMA, SARIMA, Prophet, Skincare SalesAbstract
The skincare industry has experienced rapid growth driven by increasing consumer awareness of skin health and the influence of digital trends. One of the main challenges is sales fluctuation caused by seasonal factors, promotions, and social trends. To address this issue, this study compares three time-series forecasting methods: ARIMA, SARIMA, and Prophet, using monthly skincare sales data from 2023. The research process involved data cleaning, stationarity testing, trend-seasonal decomposition, and model development with evaluation metrics including RMSE, MAE, and MASE. The results indicate that Prophet outperforms the other models in capturing long-term trends and irregular seasonal patterns, while ARIMA performs better on stable data without strong seasonality, and SARIMA is effective only with carefully optimized seasonal parameters. These findings highlight Prophet as the most accurate and practical approach for forecasting monthly skincare sales. This study is expected to provide practical contributions for skincare companies in managing inventory, planning production, and developing data-driven marketing strategies.
References
[1] S. Kwarteng and P. Andreevich, “Comparative Analysis of ARIMA, SARIMA and Prophet Model in Forecasting,” Res. Dev., vol. 5, no. 4, pp. 110–120, 2024, doi: 10.11648/j.rd.20240504.13.
[2] Y. Ensafi, S. H. Amin, G. Zhang, and B. Shah, “Time-series forecasting of seasonal items sales using machine learning – A comparative analysis,” Int. J. Inf. Manag. Data Insights, vol. 2, no. 1, p. 100058, 2022, doi: 10.1016/j.jjimei.2022.100058.
[3] V. Kramar and V. Alchakov, “Time-Series Forecasting of Seasonal Data Using Machine learning Methods,” Algorithms, vol. 16, no. 5, 2023, doi: 10.3390/a16050248.
[4] V. N. Aziza, F. H. Moh’d, F. A. Maghfiroh, K. A. Notodiputro, and Y. Angraini, “Performance Comparison of Sarima Intervention and Prophet Models for Forecasting the Number of Airline Passenger At Soekarno-Hatta International Airport,” Barekeng, vol. 17, no. 4, pp. 2107–2120, 2023, doi: 10.30598/barekengvol17iss4pp2107-2120.
[5] B. M. Osman and A. M. S. Muse, “Predictive analysis of Somalia’s economic indicators using advanced machine learning models,” Cogent Econ. Financ., vol. 12, no. 1, p., 2024, doi: 10.1080/23322039.2024.2426535.
[6] A. Avinash, A. Widjaja, and O. Karnalim, “Analisis Perbandingan Algoritma Machine learning untuk Forecasting Persediaan Produk Barang Pokok,” J. Tek. Inform. dan Sist. Inf., vol. 10, no. 2, pp. 361–378, 2024, doi: 10.28932/jutisi.v10i2.9357.
[7] K. Sharma, R. Bhalla, and G. Ganesan, “Time Series Forecasting Using FB-Prophet,” CEUR Workshop Proc., vol. 3445, pp. 59–65, 2022.
[8] R. Ospina, J. A. M. Gondim, V. Leiva, and C. Castro, “An Overview of Forecast Analysis with ARIMA Models during the COVID-19 Pandemic: Methodology and Case Study in Brazil,” Mathematics, vol. 11, no. 14, pp. 1–18, 2023, doi: 10.3390/math11143069.
[9] Y. Zhao, “Journal of Innovation and Development E-commerce Demand Forecasting Using SARIMA Model and K-means Clustering Analysis,” J. Innov. Dev., vol. 7, no. 1, pp. 1–6, 2024.
[10] J. Zhao and C. Zhang, “Research on sales forecast based on prophet-SARIMA combination model,” J. Phys. Conf. Ser., vol. 1616, no. 1, 2020, doi: 10.1088/1742-6596/1616/1/012069.
