Application of the LSTM Model for Demand Forecasting in Plastic Packaging Production

Year 2025, Volume: 2 Issue: 2, 138 - 155, 27.11.2025

Merve Yilmaz , Alptekin Durmuşoğlu

Abstract

Under increasing competitive pressures and ever-changing global conditions, companies are developing strategies to forecast future trends in order to ensure their long-term sustainability. The most critical tools in these strategies are short- and long-term demand forecasts. Accurate demand forecasting provides a valuable opportunity to minimize future revenue losses, which largely depends on the effective operation of the forecasting system. Based on the forecast results, management plans for future activities are developed, and operating costs are optimized. The plastic packaging industry is highly dynamic and is particularly sensitive to demand uncertainties due to intense competition, changing consumer behavior, and fluctuations in raw material prices. In this context, the accuracy of demand forecasts is critical for the efficient planning of production processes. In this study, the Long Short-Term Memory (LSTM) algorithm, a machine learning-based approach rooted in time series analysis, is used to forecast demand in the plastic packaging industry. Sales volumes of a plastic packaging company were collected every 15 days between January 2015 and December 2024, and 70% of these 240 data points were separated as training data, while 30% were used as test data. Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) metrics were used to evaluate the suitability and performance of the LSTM model for demand forecasting in plastic packaging production. Monthly sales volumes for 2025 were calculated using the forecast model. This research will make significant contributions to critical areas such as optimizing production processes, reducing inventory costs, and increasing customer satisfaction.

Keywords

Plastic packaging , Demand forecasting , LSTM

Project Number

1

References

• Abbasimehr, H., and Paki, R. (2021). Improving time series forecasting using LSTM and attention models. Journal of Ambient Intelligence and Humanized Computing, 13(1), 673 691.
• Akçay, H., and Yiltas-Kaplan, D. (2024). Zaman serileri tahminleme algoritmaları ile kontör tüketim tahminlemesi ve karşılaştırmalı uygulaması. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi, 27(1), 166–189.
• Albeladi, K., Zafar, B., and Mueen, A. (2023). Time series forecasting using LSTM and ARIMA. International Journal of Advanced Computer Science and Applications, 14(1), 313-320.
• Anwar, M. A., Suprihatin, S., Sasongko, N. A., Najib, M., and Pranato, B. (2024). Challenges and prospects of multilayer plastic waste management in several countries: A systematic literature review. Case Studies in Chemical and Environmental Engineering, 10, 100911.
• Aytaç, U. C., Küçükyılmaz, T., and Tarakçıoğlu, G. S. (2022). Comparison of time series models for predicting online gaming company revenue. Journal of Statistics & Applied Science, (6), 26–36.
• Bachu, Y. (2019). Packaging demand forecasting in logistics using deep neural networks [Master’s thesis, Blekinge Institute of Technology]. DiVA Portal.
• Çalık, S. (2025). Core network anomaly detectıon usıng the Lstm model and comparıson wıth varıous unsupervısed learnıng methods [Master's thesis, İstanbul Teknik Üniversitesi]. Büyük Veri ve İş Analitiği Ana Bilim Dalı.
• Dinata, M. C., and Suharjito, S. (2024). Materials inventory optimization using various forecasting techniques and purchasing quantity in packaging industry. Journal of Industrial Engineering and Management, 17(2), 321–343.
• Erdal, G. (2020). İlaç sanayinde kullanılan plastik ambalajlar. Uluslararası Sanat Tasarım ve Eğitim Dergisi (USTAD), 5(1), 25-33.
• Eraslan, İ. H., Karataş, A., and Kaya, H. (2007). Türk plastik sektörünün rekabetçilik analizi. İstanbul Ticaret Üniversitesi Sosyal Bilimler Dergisi, 6(11), 203-219.
• Fatima, S. S. W., and Rahimi, A. (2024). A review of time-series forecasting algorithms for industrial manufacturing systems. Machines, 12, 380.
• Gür, Y. E., & Eşidir, K. A. (2023). Yapay sinir ağları ile Türkiye plastik sektörü ithalat tahmini: 2023 yılı nisan-aralık ayları. Akademik Hassasiyetler, 10(23), 91–114.
• Işık, D. N. (2023). Zaman serileri tahmininde melez bir yaklaşım [Yüksek lisans tezi, Fatih Sultan Mehmet Vakıf Üniversitesi]. Bilgisayar Mühendisliği Ana Bilim Dalı.
• Kalinggo, B. A., and Zulkarnain. (2020). Time series forecasting for non-stationary data: A case study of petrochemical product price. In Proceedings of the 3rd Asia Pacific Conference on Research in Industrial and Systems Engineering, APCORISE '20 (86–92).
• Keskin, B., and Koçoğlu, Ş. (2022). Ambalaj sektörünün sürdürülebilirliği ve petrol bazlı plastik: Plastik ambalaj sektörünün petrole olan bağımlılığının analizi. Avrupa Bilim ve Teknoloji Dergisi, 35, 252-258.
• Keskin, F. D., Çiçekli, U. G., and İçli, İ. D. (2022). Prediction of failure categories in plastic extrusion process with deep learning. Journal of Intelligent Systems: Theory and Applications, 5(1), 27–34.
• Khokhar, A., Talpur, S., and Memon, M. A. (2023). Comparative analysis of LSTM, BILSTM, and ARIMA for time series forecasting on 116 years of temperature and rainfall data from Pakistan. International Journal of Scientific Research in Computer Science, Engineering and Information Technology (IJSRCSEIT), 10(2), 350–357.
• Leong, L. A. (2023). Forecasting SOXX with long short-term memory (LSTM) neural networks based on varying sampling periods [Doctoral dissertation, Selinus University]. The Department of Finance & Economics. Lewis, C. D. (1982). Industrial and business forecasting methods. Butterworths Publishing. London, 40.
• Malashin, I., Tynchenko, V., Gantimurov, A., Nelyub, V., and Borodulin, A. (2024). Applications of long short-term memory (LSTM) networks in polymeric sciences: A review. Polymers (Basel), 16(18), 2607.
• Rahimov, S. (2024). Derin öğrenme ile borsa fiyatlarının tahmini: LSTM yaklaşımı [Master's thesis, Ankara Üniversitesi]. Ankara Üniversitesi Enstitüsü.
• Uçal, B. (2022). Plastik Ambalaj Sektör Raporu: Ocak – Aralık 2022. İstanbul Kimyevi Maddeler ve Mamulleri İhracatçıları Birliği (İKMİB).
• Yamak, P. T., Yujian, L., and Gadosey, P. K. (2019). A comparison between ARIMA, LSTM, and GRU for time series forecasting. In Proceedings of the 2019 2nd International Conference on Algorithms, Computing and Artificial Intelligence (ACAI'19) (49–55).
• Yan, S. (2016). Understanding LSTM and its diagrams. Medium. https://medium.com/mlreview/understanding-lstm-and-its-diagrams-37e2f46f1714.
• Le, X.-H., Ho, H. V., Lee, G., and Jung, S. (2019). Application of Long Short-Term Memory (LSTM) Neural Network for Flood Forecasting. Water Resources Management, 33(12), 4151-4163.
• Tüler, A. (2025). Implementatıon Of Long Short-Term Memory (LSTM) Deep Learnıng Algorıthm To Predıct Flow Rate And Water Qualıty Index (WQI) [Master's thesis, İzmir Ekonomi Üniversitesi]. Bilgisayar Mühendisliği Bilim Dalı.
• Wang, C. C., Chang, H. T., and Chien, C. H. (2022). Hybrid LSTM-ARMA demand-forecasting model based on error compensation for integrated circuit tray manufacturing. Mathematics, 10(13), 2158.
• Witt, S. F., and Witt, C. A. (1992). Modeling and forecasting demand in tourism. Academic Press.