血液是醫療行為上不可或缺的重要資源，由於血液屬於一種固定壽命之易腐性存貨，因此血庫管控必須適當，以避免血液庫存量的不足導致醫療風險或存貨過多造成的社會資源浪費。在所有血液製品中，血小板的保存壽命最為短暫，自離開捐贈者體內後僅剩餘約五天壽命，故除了捐贈取得方式相對複雜之外，更是保存不易之血品，若負責管理血液的區域血液中心沒有建立適當的存貨管理策略，將會造成血小板存貨短缺及過期的情況發生，以本研究之研究對象為例，現行的血小板存貨管理方式仍有可改善之處，若能使用準確性更高的預測方式，並建立完善的存貨策略，將可有效的減少血小板的短缺及過期發生，進而使存貨相關成本最小化。而過去針對血小板存貨管理問題的相關研究，皆使用常見的存貨策略進行存貨管理，但常見的存貨策略像是(s, Q)策略或目標存貨水準策略，前者將會訂購固定存貨量，後者則是訂購一定存貨量到固定存貨水平，使得這些存貨策略無法反映出真實需求的變動或趨勢性，進而導致存貨管理的判斷不夠準確，增加無謂的存貨成本。
因此，本研究將建立以長短期記憶網路搭配基因演算法輔助尋找時窗大小之預測模型，並設計依據預測供應量變動之(ss, D)存貨策略，透過預測模型以解決血液中心現行預測方法之問題，藉由(ss, D)存貨策略達成可反映真實需求變動情形之目的，以獲得最佳存貨策略參數，並利用實際血液中心之血小板供應資料進行本研究模型之驗證。經由與現行管理方法比較後發現，於預測方法方面，本研究提出之預測模型具有較高準確度，於存貨策略方面，加入常見的存貨策略一併比較後，可知本研究所提出之(ss, D)策略能獲得較低的存貨成本，並且不會發生短缺及過期情形，此外，透過敏感度分析可知，安全庫存對於(ss, D)策略有直接的影響，當安全庫存位於一特定值時，將能獲得最小化總存貨成本並且無短缺及過期情形發生。

Blood is a perishable stock with a limited storage life. Among all blood products, platelets have the shortest shelf life with five days after leaving the human body. Therefore, platelets are not only complicated in donation and obtaining process, but also relatively challenging to store. If a regional blood center fails to establish appropriate inventory strategies, it could lead to platelet shortage and expiration. Blood is also a crucial resource to medical care. Hence, appropriate blood bank management is critical to avoid medical risks caused by blood shortage or social resource wastes due to excessive blood inventory.
There is still room for improvement of the current platelet inventory management. The occurrence of shortage and expiration of platelet inventory could be effectively reduced if we can adopt methods with higher accuracy and build up comprehensive inventory management strategies. In the past, related research on platelet inventory management is to leverage traditional inventory policies for inventory management. However, those inventory policies such as (s, Q) policy or order-up-to policy cannot respond the changes or trends of real demand in inventory and could lead to inaccurate judgments on inventory management and increase unnecessary inventory costs. Therefore, this research aims at establishing a predictive model based on the Long Short-term memory (LSTM). We also use the Genetic Algorithm (GA) to find the optimal parameters. In addition, we design the (ss, D) inventory policy based on the predicted supply quantity change to solve the problems of the current prediction solution from the blood center to achieve the goal of reflecting demand changes precisely with (ss, D) policy, and successfully verify the model in this study with actual supply data of platelets. After comparing to the current inventory management methods, in the aspect of prediction method, it is found that the predictive model in this study produces a prediction method of higher accuracy. For inventory management, after comparing altogether with common inventory policies, it is noticed that the (ss, D) policy in this study generate lower inventory costs, and shortage and expiration are no longer existed.

黃誼昀(民 109)。區域血液中心之存貨政策分析。國立成功大學工業與資訊管理研究所碩士論文。
Abbasi, B., Babaei, T., Hosseinifard, Z., Smith-Miles, K., & Dehghani, M. (2020). Predicting solutions of large-scale optimization problems via machine learning: A case study in blood supply chain management. Computers & Operations Research, 119.
Alshwaheen, T. I., Hau, Y. W., Ass’Ad, N., & Abualsamen, M. M. (2020). A Novel and Reliable Framework of Patient Deterioration Prediction in Intensive Care Unit Based on Long Short-Term Memory-Recurrent Neural Network. IEEE Access, 9, 3894-3918.
Baesler, F., Nemeth, M., Martinez, C., & Bastias, A. (2014). Analysis of inventory strategies for blood components in a regional blood center using process simulation. Transfusion, 54(2), 323-330.
Belien, J., & Force, H. (2012). Supply chain management of blood products: A literature review. European Journal of Operational Research, 217(1), 1-16.
Bermúdez, J., Achanccaray, P., Sanches, I., Cue, L., Happ, P., & Feitosa, R. (2017). Evaluation of recurrent neural networks for crop recognition from multitemporal remote sensing images. In Anais do XXVII Congresso Brasileiro de Cartografia. (pp.800-804)
Blake, J. T., Thompson, S., Smith, S., Anderson, D., Arellana, R., & Bernard, D. (2003). Optimizing the platelet supply chain in Nova Scotia. Paper presented at the Proceedings of the 29th Meeting of the European Working Group on Operational Research Applied to Health Services(ORAHS). (pp.47-66)
Cohen, M. A., & Pierskalla, W. P. (1979). Target inventory levels for a hospital blood-bank or a decentralized regional blood banking system. Transfusion, 19(4), 444-454.
Dianawati, F., Surjandari, I., & Nafitri, R. (2012). Forecasting methods for determining the level of safety stock in electronic industry. In Industrial Engineering: Innovative Networks (pp. 359-366).
Duan, Q. L., & Liao, T. W. (2013). A new age-based replenishment policy for supply chain inventory optimization of highly perishable products. International Journal of Production Economics, 145(2), 658-671.
Heddle, N. M., Liu, Y., Barty, R., Webert, K. E., Whittaker, S., Gagliardi, K., Lauzon, D., Owens, W. (2009). Factors affecting the frequency of red blood cell outdates: an approach to establish benchmarking targets. Transfusion, 49(2), 219-226.
Hemmelmayr, V., Doerner, K. F., Hartl, R. F., & Savelsbergh, M. W. P. (2009). Delivery strategies for blood products supplies. OR Spectrum, 31(4), 707-725.
Hinton, G. E., Osindero, S., & Teh, Y. W. (2006). A fast learning algorithm for deep belief nets. Neural Computation, 18(7), 1527-1554.
Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735-1780.
Jennings, J. B. (1973). Blood bank inventory control. Management Science Series B-Application, 19(6), 637-645.
Kamp, C., Heiden, M., Henseler, O., & Seitz, R. (2010). Management of blood supplies during an influenza pandemic. Transfusion, 50(1), 231-239.
Kantasa-ard, A., Nouiri, M., Bekrar, A., Ait el cadi, A., & Sallez, Y. (2020). Machine learning for demand forecasting in the physical internet: a case study of agricultural products in Thailand. International Journal of Production Research, 1-25.
Kazemi, S. M., Rabbani, M., Tavakkoli-Moghaddam, R., & Shahreza, F. A. (2017). Blood inventory-routing problem under uncertainty. Journal of Intelligent & Fuzzy Systems, 32(1), 467-481.
Khaldi, R., El Afia, A., Chiheb, R., & Faizi, R. (2017). Artificial neural network based approach for blood demand forecasting: Fez transfusion blood center case study. In Proceedings of the 2nd international Conference on Big Data, Cloud and Applications.
Morish, M., Ayob, Y., Noris Naim, H. S., Muhamad, N. A., & Yusoff, N. M. (2012). Quality indicators for discarding blood in the National Blood Center, Kuala Lumpur. Asian Journal of Transfusion Science, 6(1), 19.
Nahmias, S. (1982). Perishable inventory-theory - a review. Operations Research, 30(4), 680-708.
Osorio, A. F., Brailsford, S. C., & Smith, H. K. (2015). A structured review of quantitative models in the blood supply chain: a taxonomic framework for decision-making. International Journal of Production Research, 53(24), 7191-7212.
Osorio, A. F., Brailsford, S. C., Smith, H. K., Forero-Matiz, S. P., & Camacho-Rodriguez, B. A. (2017). Simulation-optimization model for production planning in the blood supply chain. Health Care Management Science, 20(4), 548-564.
Pereira, A. (2004). Performance of time-series methods in forecasting the demand for red blood cell transfusion. Transfusion, 44(5), 739-746.
Perera, G., Hyam, C., Taylor, C., & Chapman, J. F. (2009). Hospital Blood Inventory Practice: the factors affecting stock level and wastage. Transfusion Medicine, 19(2), 99-104.
Pierskalla, W. P. (2005). Supply chain management of blood banks. In Operations Research and Health Care, 70, 103-145.
Prastacos, G. P. (1984). Blood inventory management - an overview of theory and practice. Management Science, 30(7), 777-800.
Raafat, F. (1991). Survey of literature on continuously deteriorating inventory models. Journal of the Operational Research Society, 42(1), 27-37.
Rajendran, S., & Ravindran, A. R. (2017). Platelet ordering policies at hospitals using stochastic integer programming model and heuristic approaches to reduce wastage. Computers & Industrial Engineering, 110, 151-164.
Rajendran, S., & Ravindran, A. R. (2019). Inventory management of platelets along blood supply chain to minimize wastage and shortage. Computers & Industrial Engineering, 130, 714-730.
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986). Learning representations by back-propagating errors. Nature, 323(6088), 533-536.
Shahid, F., Zameer, A., & Muneeb, M. (2021). A novel genetic LSTM model for wind power forecast. Energy, 223, 120069.
Shih, H., & Rajendran, S. (2019). Comparison of Time Series Methods and Machine Learning Algorithms for Forecasting Taiwan Blood Services Foundation's Blood Supply. Journal of Healthcare Engineering, 2019.
Stajkowski, S., Kumar, D., Samui, P., Bonakdari, H., & Gharabaghi, B. (2020). Genetic-Algorithm-Optimized Sequential Model for Water Temperature Prediction. Sustainability, 12(13), 5374.
Stanger, S. H. W., Yates, N., Wilding, R., & Cotton, S. (2012). Blood Inventory Management: Hospital Best Practice. Transfusion Medicine Reviews, 26(2), 153-163.
van Donselaar, K., van Woensel, T., Broekmeulen, R., & Fransoo, J. (2006). Inventory control of perishables in supermarkets. International Journal of Production Economics, 104(2), 462-472.
Wee, H. M. (1993). Economic production lot-size model for deteriorating items with partial back-ordering. Computers & Industrial Engineering, 24(3), 449-458.
Zhao, P., & Liu, J. (2010, August). The product safety stock prediction method based on artificial neural network. In 2010 International Conference of Information Science and Management Engineering (Vol. 1, pp. 299-302). IEEE.
Zhou, D. M., Leung, L. C., & Pierskalla, W. P. (2011). Inventory Management of Platelets in Hospitals: Optimal Inventory Policy for Perishable Products with Regular and Optional Expedited Replenishments. M&SOM-Manufacturing & Service Operations Management, 13(4), 420-438.