數位科技的普及加速了市場創新和傳播速度，使品牌能夠透過社交媒體、電商平台和數據分析工具更直接地與消費者互動。消費者趨向於追求多樣性和個性化選擇，並更注重品質、可持續性和時尚性。時尚製造產業不具備傳統生命週期特徵，其無特定流行趨勢和使用期限的特性使得傳統預測方法難以適用。需求不確定性使得準確的銷售預測至關重要，探索有效的需求預測方法有助於提高預測精確度，應對市場變化，降低庫存風險，並提升生產及管理效率，輔助管理者做出應對的決策。
本研究以一家安全帽製造公司為實證分析對象，建構股市指標應用於需求需求數據之創新的預測模式，包含MACD和KD指標，傳統預測法包含移動平均法、指數平滑法，機器學習中包含ARIMA、決策樹迴歸、迴歸支援向量機、XGBoost Regressor等七種預測模型，建立具有信度以及效度之預測模式。首先，對所蒐集之數據資料進行資料前處理(Data preprocessing)，再藉由網格搜索法(Grid Search)結合交叉驗證(Cross-validation)找出各預測模型之最佳參數組合。最後比較預測模型於各情境之表現，並選擇各情境之最佳預測模型。
實驗結果表明，低、中、高變異類別在原始數據下的最佳預測模型為XGBoost模型，在需求情境下最佳預測模型為四期移動平均法以及XGBoost模型。實驗平均絕對百分比誤差皆低於30%，對於需求波動大的相關產業，已是足夠輔助判斷決策的一個標準依據，達成本研究建構可靠的需求預測模型之目的。

The widespread adoption of digital technology has accelerated market innovation and communication speed, enabling brands to interact more directly with consumers through social media, e-commerce platforms, and data analytics tools. Consumers increasingly seek diversity and personalized choices, placing greater emphasis on quality, sustainability, and fashionability. The fashion manufacturing industry lacks traditional lifecycle characteristics, making conventional forecasting methods less applicable. The uncertainty of demand underscores the importance of accurate sales forecasting, exploring effective demand prediction methods to enhance forecast accuracy, adapt to market changes, reduce inventory risks, and improve production and management efficiency, aiding managers in making informed decisions.
This study focuses on a safety helmet manufacturing company as the empirical analysis subject, constructing innovative predictive models using stock market indicators applied to demand data, including MACD and KD indicators. Traditional forecasting methods encompass moving averages and exponential smoothing, while machine learning includes ARIMA, regression decision trees, regression support vector machines, and XGBoost Regressor, among others, totaling seven prediction models. A reliable and valid predictive model is established by preprocessing the collected data and identifying the best parameter combinations for each model through grid search combined with cross-validation. Finally, the performance of the predictive models in various scenarios is compared, selecting the best predictive model for each scenario.
The experimental results indicate that the XGBoost model is the best predictive model for low, medium, and high variability categories in the original data, while the four-period moving average method and XGBoost model are the best predictive models in the demand scenario. The average absolute percentage errors in the experiments are all below 30%, which is considered a sufficient standard for decision-making assistance in industries with high demand volatility, achieving the goal of constructing reliable demand forecasting models in this study.

山口雄大 (2023)，《驚人的AI需求預測》，商周出版，台北市。
Appel, G. (1979). The Moving Average Convergence-divergence Trading Method Advanced Version, New York, United States.
Ali, M.M., & Boylan, J.E. (2012). On the Effect of Non-optimal Forecasting Methods on Supply Chain Downstream Demand. The Institute of Mathematics and its Applications Journal of Management Mathematics, 23(1), 81-98.
Anghel, G. D. I. (2015). Stock market efficiency and the MACD. Evidence from countries around the world. Procedia Economics and Finance, 32, 1414-1431.
Alawad, W., Zohdy, M., & Debnath, D. (2018). Tuning hyperparameters of decision tree classifiers using computationally efficient schemes. IEEE First International Conference on Artificial Intelligence and Knowledge Engineering, September 26-28, Laguna Hills, California, United States.
Azizi, A., Rooki, R., & Mollayi, N. (2020). Modeling and prediction of wear rate of grinding media in mineral processing industry using multiple kernel support vector machine. Springer Nature Applied Sciences, 2(9), 1469.
Ayala, J., García-Torres, M., Noguera, J. L. V., Gómez-Vela, F., & Divina, F. (2021). Technical analysis strategy optimization using a machine learning approach in stock market indices. Knowledge Based Systems, 225, 107119.
Abbasimehr, H., Paki, R., & Bahrini, A. (2023). A novel XGBoost-based featurization approach to forecast renewable energy consumption with deep learning models. Sustainable Computing Informatics and Systems, 38, 100863.
Brown, R. G. (1959). Statistical Forecasting for Inventory Control, McGraw Hill, New York, United States.
Bass FM. (1969). A new product growth for model consumer durables. Management Sciences, 15(5), 215-227.
Browne, J., Harhen, J., & Shivnan, J. (1996). Production Management Systems: An Integrated Perspective, Addison Wesley, Aberdeen, United Kingdom.
Breiman, L., Friedman, J., Stone, C. J., & Olshen, R. A. (1984). Classification and Regression Trees, Chapman & Hall, New York, United States.
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5-32.
Berk, R. A., (2008). Classification and regression trees (CART), Available: https://link.springer.com/content/pdf/10.1007/978-0-387-77501-2_3.pdf (2024/03/15 取得)
Bergstra, J., & Bengio, Y. (2012). Random search for hyper-parameter optimization. Journal of Machine Learning Research, 13(2), 281-305.
Babai, M. Z., Ali, M. M., Boylan, J. E., & Syntetos, A. A. (2013). Forecasting and inventory performance in a two-stage supply chain with ARIMA (0, 1, 1) demand: Theory and empirical analysis. International Journal of Production Economics, 143(2), 463-471.
Cox W. E. (1967). Product life cycles as marketing models. J. Bus, 40(4), 375-384.
Chong, T. T. L., & Ng, W. K. (2008). Technical analysis and the London stock exchange: testing the MACD and RSI rules using the FT30. Applied Economics Letters, 15(14), 1111-1114.
Da Veiga, C. P., Da Veiga, C. R. P., Catapan, A., Tortato, U., & Da Silva, W. V. (2014). Demand forecasting in food retail: A comparison between the Holt-Winters and ARIMA models. World Scientific and Engineering Academy and Society Transactions on Business and Economics, 11(1), 608-614.
Drucker, H., Burges, C. J., Kaufman, L., Smola, A., & Vapnik, V. (1996). Support vector regression machines. Advances in Neural Information Processing Systems, 9, 155-161.
Elattar, E. E., Goulermas, J., & Wu, Q. H. (2010). Electric load forecasting based on locally weighted support vector regression. IEEE Transactions on Systems, Man and Cybernetics, Part C (Applications and Reviews), 40(4), 438-447.
Faraway, J. J. (2016). Does data splitting improve prediction. Statistics and Computing, 26, 49-60.
Fildes, R., Ma, S., & Kolassa, S. (2022). Retail forecasting: Research and practice. International Journal of Forecasting, 38(4), 1283-1318.
Fang, Z. G., Yang, S. Q., Lv, C. X., An, S. Y., & Wu, W. (2022). Application of a data-driven XGBoost model for the prediction of COVID-19 in the United States: a time-series study. The British Medical Journal Open, 12(7), 56685.
Gardner Jr, E. S. (1985). Exponential smoothing: The state of the art. International Journal of Forecasting, 4(1), 1-28.
Gallien J, Mersereau AJ, Garro A, Mora AD, Vidal MN. (2015). Initial shipment decisions for new products at Zara. Operational Research, 63(2), 269-286.
Headen RS. (1966). The Introductory Phases of the Life Cycle for New Grocery Products: Consumer Acceptance and Competitive Behavior, Ph.D. Business Administration, Harvard University, Cambridge, Massachusetts, United States.
Hopp, W. J., & Spearman, M. L. (2011). Factory Physics, Waveland Press, Long Grove, United States.
Hsu, C. W., Chang, C. C., & Lin, C. J. (2003). A Practical Guide to Support Vector Classification, Available: https://www.csie.ntu.edu.tw/~cjlin/papers/guide/guide.pdf (2024/03/21 取得)
Holt, C. C. (2004). Forecasting seasonal and trends by exponentially weighted moving averages. International Journal of Forecasting, 20(1), 5-10.
Huang, H., Wei, X., & Zhou, Y. (2022). An overview on twin support vector regression. Neurocomputing, 490, 80-92.
Lewis, C. D. (1982). Industrial and Business Forecasting Methods: A Practical Guide to Exponential Smoothing and Curve Fitting, Butterworth Scientific, London.
Lane, G. C. (1984). Lane’s stochastics. Technical Analysis of Stocks and Commodities, 2(3), 80.
Lee, H. L., Padmanabhan, V., and Whang, S. (1997). Information distortion in a supply chain: The bullwhip effect. Management Science, 43(4), 546-558.
McGill, R., Tukey, J. W., & Larsen, W. A. (1978). Variations of box plots. The American Statistician, 32(1), 12-16.
Minner, S., & Kiesmüller, G. P. (2012). Dynamic product acquisition in closed loop supply chains. International Journal of Production Research, 50(11), 2836-2851.
Makridakis, S., Wheelwright, S. C., & Hyndman, R. J. (2008). Forecasting Methods and Applications, John Wiley & Sons, United States.
Mittal, V., & Schaposnik, L. P. (2023). Housing market forecasts via stock market indicators. Heliyon, 9(5), 16286.
Nenni, M. E., Giustiniano, L., & Pirolo, L. (2013). Demand forecasting in the fashion industry: a review. International Journal of Engineering Business Management, 5, 37.
Noorunnahar, M., Chowdhury, A. H., & Mila, F. A. (2023). A tree based eXtreme Gradient Boosting (XGBoost) machine learning model to forecast the annual rice production in Bangladesh. Public Library of Science, 18(3), 283452.
Ostertagova, E., & Ostertag, O. (2012). Forecasting using simple exponential smoothing method. Acta Electrotechnica et Informatica, 12(3), 62.
Pongdatu, G. A. N., & Putra, Y. H. (2018). Seasonal time series forecasting using SARIMA and Holt Winter’s exponential smoothing. In Institute of Physics Conference Series: Materials Science and Engineering, 407(1), 012153, May 9, Bandung, Indonesia.
Pemathilake, R. G. H., Karunathilake, S. P., Shamal, J. L. A. J., & Ganegoda, G. U. (2018). Sales forecasting based on autoregressive integrated moving average and recurrent neural network hybrid model. In 2018 14th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery, 27-33, July 28-30, Huangshan, China.
Rink, D. R., & Swan, J. E. (1979). Product life cycle research: A literature review. Journal of Business Research, 7(3), 219-242.
Rego, J. R., and Mesquita, M. A. (2015). Demand forecasting and inventory control: A simulation study on automotive spare parts. International Journal of Production Economics, 161, 1-16.
Ray, S. (2019). A quick review of machine learning algorithms. In 2019 International Conference on Machine Learning, Big Data, Cloud and Parallel Computing, 35-39, February 14-16, Faridabad, India.
Ryu, S. E., Shin, D. H., & Chung, K. (2020). Prediction model of dementia risk based on XGBoost using derived variable extraction and hyper parameter optimization. IEEE Access, 8, 177708-177720.
Sani, B., & B. G. Kingsman. (1997). Selecting the Best Periodic Inventory Control and Demand Forecasting Methods for Low Demand Items. The Journal of the Operational Research Society, 48(7), 700-713.
Syntetos, A. A., & J. E. Boylan. (2006). On the Stock Control Performance of Intermittent Demand Estimators. International Journal of Production Economics, 103(1), 36-47.
Song, Y. Y., & Ying, L. U. (2015). Decision tree methods: applications for classification and prediction. Shanghai Archives of Psychiatry, 27(2), 130.
Shokri, S., Sadeghi, M. T., Marvast, M. A., & Narasimhan, S. (2015). Improvement of the prediction performance of a soft sensor model based on support vector regression for production of ultra-low sulfur diesel. Petroleum Science, 12, 177-188.
Smyl, S. (2020). A hybrid method of exponential smoothing and recurrent neural networks for time series forecasting. International Journal of Forecasting, 36(1), 75-85.
Townsend, J. T. (1971). Theoretical analysis of an alphabetic confusion matrix. Perception & Psychophysics, 9, 40-50.
Vernon, R. (1966). International investment and international trade in the product cycle . Quarterly Journal of Economics , 80(2), 90-207.
Varghese, V., & Rossetti, M. (2008). A parametric bootstrapping approach to forecast intermittent demand. Proceedings of the 2008 Industrial Engineering Research Conference, 857, December 19-21, Washington, D.C., United States.
Winters, P. R. (1960). Forecasting sales by exponentially weighted moving averages. Management science, 6(3), 324-342.
Wang, Y., Pan, Z., Zheng, J., Qian, L., & Li, M. (2019). A hybrid ensemble method for pulsar candidate classification. Astrophysics and Space Science, 364, 1-13.
Wu, B. (2023). Prediction of the Number of Wordle Report Results Based on ARIMA and BIC. In 2023 IEEE 3rd International Conference on Data Science and Computer Application (ICDSCA), 606-611, October 27-29, Dalian, China.
Xie, Y., Jin, M., Zou, Z., Xu, G., Feng, D., Liu, W., & Long, D. (2020). Real-time prediction of docker container resource load based on a hybrid model of ARIMA and triple exponential smoothing. IEEE Transactions on Cloud Computing, 10(2), 1386-1401.
Yu, X., Qi, Z., & Zhao, Y. (2013). Support vector regression for newspaper magazine sales forecasting. Procedia Computer Science, 17, 1055-1062.