Demand Forecasting For Furniture Industry With Multi-Variable Time Series Models

Öz

Demand forecasting is important for businesses in order to be prepared for any adverse situations that may arise in the future. Time series models are a useful and reliable tool for demand forecasting. These models enable businesses to improve planning and decision-making by providing more accurate and reliable demand forecasts. This study aims to forecast the monthly door sales quantities of a business in the furniture industry using various multi-variable Time Series methods. The variable set includes dollar exchange rate, inflation, and interest rates. Models considered and compared are ARIMA-SARIMAX, Multiple Linear Regression and Holt-Winters. Analysis are conducted using real-life sales data obtained for years 2005 to 2019. Required data on the dollar exchange rate, inflation, and interest rates are obtained from the Central Bank of Türkiye. By incorporating these variables, the study aims to enhance the accuracy and reliability of the predictions, providing valuable insights for the industry. Results show that ARIMA-SARIMAX produce more accurate forecasts compared to (Holt-Winters and Multiple Linear Regression. It is observed that the time series model that takes into account the seasonality and trend factors provide better forecast results in the furniture industry. The findings highlight the importance of advanced forecasting methods in maintaining competitive advantage. This approach not only supports strategic planning but also helps in optimizing inventory levels.

Anahtar Kelimeler

• Furniture Sector
• Demand Forecasting
• Holt-Winters
• ARIMA-SARIMAX
• Multiple Linear Regression

Çok Değişkenli Zaman Serisi Modelleriyle Mobilya Endüstrisi İçin Talep Tahmini

Öz

İşletmelerin gelecekte oluşacak herhangi bir olumsuz duruma hazırlıklı olmaları için talep tahmini çalışmalarına önem vermeleri gerekir. Zaman serileri modeli, mobilya sektöründe talep tahmini için kullanışlı ve güvenilir bir araçtır. Zaman serileri modeli, işletmelere daha doğru ve güvenilir talep tahminleri sağlayarak, daha iyi planlama ve karar verme imkanı sunmaktadır. Bu çalışmada Zaman Serileri ile talep tahmin yöntemi kullanılarak işletmenin aylık kapı satış miktarlarının tahmini yapılması amaçlanmaktadır. Zaman serileri modelinin (ARIMA, SARIMAX), diğer talep tahmin yöntemlerine (Holt-Winters, Çoklu Doğrusal Regresyon) kıyasla daha doğru tahminler ürettiği görülmüştür. Özellikle, mevsimsellik ve trend gibi faktörleri de dikkate alan zaman serileri modeli, mobilya sektöründeki talepleri daha iyi modellemeyi başarmıştır. Bu bağlamda, Alanya’da bir fabrikanın 2005-2019 yılları arasındaki kapı satış miktarı verileri tahminleme yapmak için kullanılmıştır. Dolar kuru, enflasyon ve faiz oranı verileri Merkez Bankası’ndan alınmıştır.

Anahtar Kelimeler

• Mobilya Sektörü
• Talep Tahmini
• Holt-Winters
• ARIMA-SARIMAX
• Çoklu Doğrusal Regresyon

Kaynakça

1. [1] I. Akgul, "Time Series Analysis and Forecasting Models," Suggestion Journal, vol. 1, no. 1, pp. 52–69, 1994, doi: 10.14783/maruoneri.698511.
2. [2] F. Alharbi and D. Csala, "A Seasonal Autoregressive Integrated Moving Average with Exogenous Factors (SARIMAX) Forecasting Model-Based Time Series Approach," School of Engineering, Lancaster University, vol. 7, pp. 94, 2022.
3. [3] D. Aydin, "Demand Forecast Analysis with the Help of Artificial Neural Networks and an Application in Maritime Transportation Sector," M.S. thesis, Inst. of Social Sciences, Marmara Univ., Istanbul, Türkiye, 2012. [4] C. Chase, Demand-Driven Forecasting: A Structured Approach to Forecasting. USA: John Wiley & Sons, 2013, doi: 10.1002/9781118691861.
4. [5] J. Cheng, S. Tiwari, D. Khaled, M. Mahendru, and U. Shahzad, "Forecasting Bitcoin prices using artificial intelligence: Combination of ML, SARIMA, and Facebook Prophet models," Technological Forecasting and Social Change, vol. 198, pp. 1–15, 2024, doi: 10.1016/j.techfore.2023.122938.
5. [6] J. Fattah, L. Ezzine, Z. Aman, H. El Moussami, and A. Lachhab, "Forecasting of demand using ARIMA model," International Journal of Engineering Business Management, vol. 10, pp. 1–9, 2018, doi: 10.1177/1847979018808673.
6. [7] T. Falatouri, F. Darbanian, P. Brandtner, and C. Udokwu, "Predictive Analytics for Demand Forecasting – A Comparison of SARIMA and LSTM in Retail SCM," Procedia Computer Science, vol. 200, pp. 993–1003, 2022, doi: 10.1016/j.procs.2022.01.298.
7. [8] E. Hazır, K. K. Koç, and Ş. Esnaf, "Prediction of Turkish Furniture Sales Values with a Sample Artificial Intelligence Application," National Furniture Congress Journal, pp. 1172–1182, 2015.
8. [9] S. Imece and Ö. F. Beyca, "Demand Forecasting with Integration of Time Series and Regression Models in Pharmaceutical Industry," Demand Forecasting in Pharmaceutical Industry, vol. 34, no. 3, pp. 414–423, 2022, doi: 10.7240/jeps.1127844.

9. [10] F. R. Jacobs and R. B. Chase, Operations and Supply Management: The Core, 6th ed. New York: McGraw-Hill, 2010.
10. [11] W. Jiang, X. Wu, Y. Gong, W. Yu, and X. Zhong, "Holt–Winters smoothing enhanced by fruit fly optimization algorithm to forecast monthly electricity consumption," Energy, vol. 193, pp. 1–8, 2020, doi: 10.1016/j.energy.2019.116779.
11. [12] N. Kaya, Stock Management. Ankara: İksad Publishing House, 2020.
12. [13] M. Lin, Z. Zhang, and Y. Cao, "Forecasting Supply and Demand of the Wooden Furniture Industry in China," Forest Products Journal, vol. 69, no. 3, pp. 228–238, 2019, doi: 10.13073/FPJ-D-19-00011.
13. [14] S. Makridakis, S. C. Wheelwright, and R. J. Hyndman, Forecasting: Methods and Applications. New York: John Wiley & Sons, 1997.
14. [15] E. E. Nebati, M. Taş, and G. Ertaş, "Demand Forecasting in Electricity Consumption in Turkey: Comparison with Time Series and Regression Analysis," European Journal of Science and Technology, no. 31, pp. 348–357, 2021, doi: 10.31590/ejosat.998277.
15. [16] M. Nurtaş, Z. Zhantaev, and A. Altaibek, "Earthquake time-series forecast in Kazakhstan territory: Forecasting accuracy with SARIMAX," Procedia Computer Science, vol. 231, pp. 353–358, 2024, doi: 10.1016/j.procs.2023.12.216.
16. [17] M. Karahan, "Statistical Forecasting Methods: Application of Product Demand Forecasting with Artificial Neural Networks Method," M.S. thesis, Dept. of Business Administration, Selcuk Univ., Konya, Türkiye, 2011.
17. [18] V. Önen, "Turkey's Airline Cargo Demand Forecast Modeling and Forecasting Using ARIMA Method," Journal of Management and Economics Research, vol. 18, no. 4, pp. 29–53, 2020, doi: 10.11611/yead.677319.
18. [19] V. Önen, "Turkish Airline Passenger Demand Forecast Modeling, Forecasting and Comparison with ARIMA-ARIMAX Method," Journal of Transportation and Logistics, vol. 8, no. 2, pp. 242–273, 2023, doi: 10.26650/JTL.2023.1270944.
19. [20] E. Polat, "Sales Forecast in White Goods Industry: A Data Mining Application," M.S. thesis, Inst. of Social Sciences, Uludağ Univ., Bursa, Türkiye, 2022.
20. [21] G. Rumbe, M. Hamasha, and S. A. Mashaqbeh, "A comparison of Holts-Winter and Artificial Neural Network approach in forecasting: A case study for tent manufacturing industry," Results in Engineering, vol. 21, pp. 1–6, 2024, doi: 10.1016/j.rineng.2024.101899.
21. [22] B. Salttürk, "Estimation of Product Sales Quantities with Artificial Neural Networks: An Application in Furniture Industry," M.S. thesis, Inst. of Science, Sakarya Univ., Sakarya, Türkiye, 2022.
22. [23] R. Siddiqui, M. Azmat, S. Ahmed, and S. Kummer, "A Hybrid Demand Forecasting Model for Greater Forecasting Accuracy: The Case of the Pharmaceutical Industry," Supply Chain Forum: An International Journal, vol. 23, no. 3, pp. 1–11, 2021, doi: 10.1080/16258312.2021.1967081.
23. [24] W. V. D. Silva, C. P. D. Veiga, C. R. P. D. Veiga, A. Catapan, and U. Tortato, "Demand forecasting in food retail: a comparison between the Holt-Winters and ARIMA models," WSEAS Transactions on Business and Economics, vol. 11, pp. 608–614, 2014.
24. [25] S. Ulutürk, "Refinement Models in Future Forecasting and an Application," M.S. thesis, Inst. of Social Sciences, Istanbul Technical Univ., Istanbul, Türkiye, 1994.
25. [26] M. Valipour, "Application of ARIMA model for inflow forecasting of Moghanlo River," Journal of King Saud University – Engineering Sciences, vol. 27, no. 1, pp. 72–84, 2015.
26. [27] P. E. Yaneva and H. N. Kulina, "Furniture Market Demand Forecasting Using Machine Learning Approaches," Journal of Physics: Conference Series, vol. 2675, no. 1, pp. 1–11, 2023, doi: 10.1088/1742-6596/2675/1/012008.
27. [28] M. Yücesan, "Forecasting Monthly Sales of White Goods Using Hybrid ARIMAX and ANN Models," Atatürk University Journal of Social Sciences Institute, vol. 22, no. 4, pp. 2603–2617, 2018.
28. [29] M. Yücesan, M. Gül, and E. Çelik, "Performance comparison between ARIMAX, ANN and ARIMAX-ANN hybridization in sales forecasting for furniture industry," Drvna Industrija, vol. 69, no. 4, pp. 357–370, 2018.
29. [30] B. Yüksel, "Demand Prediction in Time Series," YBS Encyclopedia, vol. 11, no. 2, pp. 1–18, 2023.
30. [31]"Statsmodels," Accessed: Sep. 21, 2024. [Online]. Available: https://www.statsmodels.org/stable/examples/notebooks/generated/statespace_sarimax_stata.html