本研究以某地區醫院的輸液需求為例，探討應用長短期記憶網路(Long Short-Term Memory, LSTM)模型於藥品需求預測與安全庫存量設定中的可行性。透過分析2019至2024年的日輸液使用數據，進行詳細資料清理與時間序列分析，建立適用於需求波動情境的LSTM模型。研究結果顯示，LSTM模型能準確捕捉數據中的趨勢與季節性變化，其預測準確性顯著優於傳統統計方法，特別在高需求波動的情境下表現出色且穩定。根據需求預測結果，本研究設置了科學化的安全庫存策略， 成功將緊急補貨次數由原本的5次降至0次，顯著改善了醫療供應鏈的運作效率與穩定性。本研究驗證了深度學習技術在醫療資源管理中的應用潛力，並建議未來研究可納入更多環境變數及供應鏈特性，進一步提升模型的適用性與準確性，促進醫療資源分配的最佳化與患者用藥安全的保障。

This study explores the feasibility of applying Long Short-Term Memory (LSTM) models to drug demand forecasting and safety stock optimization, using the intravenous fluid demand of a regional hospital as a case study. By analyzing daily usage data from 2019 to 2024, the study involved comprehensive data cleaning and time series analysis to construct an LSTM model tailored for fluctuating demand scenarios. The results demonstrate that the LSTM model accurately captures trends and seasonal patterns in the data, significantly outperforming traditional statistical methods, particularly under high-demand volatility. Based on the demand forecasting results, a scientific safety stock strategy was developed, successfully reducing the number of emergency restocks from five to zero, thereby enhancing the operational efficiency and stability of the medical supply chain. This study validates the potential of deep learning technologies in healthcare resource management and suggests future research should incorporate additional environmental variables and supply chain characteristics to further improve the model's applicability and accuracy, ultimately optimizing resource allocation and ensuring patient medication safety.

英文文獻
1.Ali, A. K. (2011). Inventory management in pharmacy practice: A review of literature. Archives of Pharmacy Practice, 2(4), 151–156.
2.Bottou, L. (2010). Large-scale machine learning with stochastic gradient descent. In Y. Lechevallier & G. Saporta (Eds.), Proceedings of COMPSTAT'2010 (pp. 177–186). Springer.
3.Bengio, Y. (2012). Practical recommendations for gradient-based training of deep architectures. In G. Montavon, G. B. Orr, & K.-R. Müller (Eds.), Neural networks: Tricks of the trade (2nd ed., pp. 437–478). Springer.
4.Cohen, M. A., & Roth, A. V. (2016). Supply chain management in healthcare institutions: Efficiency and effectiveness in the face of uncertainty. Operations Research, 64(6), 1205–1220.
5.Chopra, S., & Meindl, P. (2019). Supply chain management: Strategy, planning, and operation (7th ed.). Pearson.
6.Dube, K., Katende-Kyenda, N. L., & Matsebula, Z. C. (2020). Pharmaceutical supply chain management and performance in public healthcare institutions. International Journal of Health Planning and Management, 35(1), 140–153.
7.Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT Press.
8.Graves, A. (2013). Generating sequences with recurrent neural networks. arXiv preprint arXiv:1308.0850.
9.Gauss, C. F. (1809). Theoria motus corporum coelestium in sectionibus conicis solem ambientium. Friedrich Perthes und I. H. Besser.
10.Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735–1780.
11.Hastie, T., Tibshirani, R., & Friedman, J. (2009). The elements of statistical learning: Data mining, inference, and prediction (2nd ed.). Springer.
12.Hadley, G., & Whitin, T. M. (1963). Analysis of inventory systems. Prentice Hall.
13.Iglewicz, B., & Hoaglin, D. C. (1993). How to detect and handle outliers. ASQC Quality Press.
14.Kingma, D. P., & Ba, J. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.
15.Kuo, S. C., Ou, H. T., & Wang, C. J. (2021). Managing medication supply chains: Lessons learned from Taiwan during the COVID-19 pandemic. Journal of Global Health, 11, 03072.
16.LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11), 2278–2324.
17.Law, A. M., & Kelton, W. D. (2015). Simulation modeling and analysis (5th ed.). McGraw-Hill Education.
18.Makridakis, S., Wheelwright, S. C., & Hyndman, R. J. (1998). Forecasting: Methods and applications (3rd ed.). Wiley.
19.Makridakis, S., Spiliotis, E., & Assimakopoulos, V. (2019). The M4 Competition: 100,000 time series and 61 forecasting methods. International Journal of Forecasting, 36(1), 54–74.
20.Mbonyinshuti, F., Nkurunziza, J., Niyobuhungiro, J., & Kayitare, E. (2024). Health supply chain forecasting: A comparison of ARIMA and LSTM time series models for demand prediction of medicines. Acta Logistica, 11(2), 269–280.
21.Nash, J. F. (1950). Equilibrium points in n-person games. Proceedings of the National Academy of Sciences, 36(1), 48–49.
22.Prechelt, L. (1998). Early stopping—But when? In G. B. Orr & K.-R. Müller (Eds.), Neural networks: Tricks of the trade (pp. 55–69). Springer.
23.Qian, N. (1999). On the momentum term in gradient descent learning algorithms. Neural Networks, 12(1), 145–151.
24.Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., & Salakhutdinov, R. (2014). Dropout: A simple way to prevent neural networks from overfitting. Journal of Machine Learning Research, 15, 1929–1958.
25.Savitzky, A., & Golay, M. J. E. (1964). Smoothing and differentiation of data by simplified least squares procedures. Analytical Chemistry, 36(8), 1627–1639.
26.Shapiro, A. (2009). Stochastic programming approach to optimization under uncertainty. Mathematical Programming, 112(1), 183–220.
27.Willmott, C. J., & Matsuura, K. (2005). Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance. Climate Research, 30(1), 79–82.
28.Yuan, Y., Zhou, H., & Liu, W. (2018). Gate mechanisms in long short-term memory (LSTM) networks: A comprehensive review. Neural Networks, 105, 35–52.
29.Yao, Y., Rosasco, L., & Caponnetto, A. (2007). On early stopping in gradient descent learning. Constructive Approximation, 26(2), 289–315.
30.Zaremba, W., Sutskever, I., & Vinyals, O. (2014). Recurrent neural network regularization. arXiv preprint arXiv:1409.2329.
 
中文文獻
1.余淑玲（2014）。醫院資材管理策略研究。醫療經營管理期刊，20（1），12–20。
2.魏慶國、王舜睦（2006）。醫療資材管理之挑戰與對策。醫療管理學報，18（2），23–29。
3.陳楚杰（2002）。醫療庫存管理的經濟訂購量模式研究（國立臺灣大學碩士論文）。
4.謝秀圓（2012）。藥品庫存策略研究：耗用類型與需求預測模型之應用。醫療管理學報，24（2），45–57。
5.劉晉豪（2024）。藥品耗用類型分類對需求預測與庫存管理之影響。醫療管理研究期刊，32（1），15–28。
6.褚志鵬、謝秀圓（2014）。醫院藥費控制策略之研究。醫療管理學報，28（2），34–49。
7.何靜宜（2008）。醫院庫存管理：策略與實務。臺北：五南圖書出版股份有限公司。
8.Chopra, S.，Meindl, P.（2019）。供應鏈管理：策略、規劃與作業。台北：新加坡：Pearson Education。
 
網路文獻
1.World Health Organization (WHO).（2016）。Addressing global shortages of medicines and vaccines。取自 https://www.who.int/medicines/areas/access/shortages/en/。
2.Hinton, G.（2012）。Lecture 6e RMSProp: Divide the gradient by a running average of its recent magnitude。Coursera: Neural Networks for Machine Learning。取自 https://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf。