快時尚風潮自二十世紀開始，顛覆傳統服飾業的商業模式，逐漸走向以上市時間短、平價以及符合時尚潮流的趨勢，而產品生命週期下降，使得零售商需要不斷地推出新產品。相較於其他傳統產業，服飾業是一個更以消費者為導向的產業，然而社群媒體的蓬勃發展，導致消費者容易受到外在因子而改變偏好，因此準確的銷售預測可以避免產品缺貨或是滯銷的風險，但是對於新產品而言，缺乏歷史銷售數據使得銷售預測成為一大難題。
本研究提出一個兩階段方法，包括輔助設計與智慧預測方法，第一階段以網路爬蟲瀏覽B2C服飾網站，下載商品名稱以及銷售數量，從商品名稱拆解商品的組成元素，再重新組合成新產品來達到輔助設計，第二階段以分類分群方法來連結元素組合與銷售數量的關聯，並使用機器學習來預測新產品的銷售需求，最後使用Google搜尋趨勢蒐集新商品的元素組成作為外部資訊指標加以調整預測值，研究結果顯示，與其他機器學習模型(隨機森林和極限梯度提升)相比，智慧需求方法可有效地降低至少45.79%的均方誤差(MSE)、至少26.35%的均方根誤差(RMSE)以及至少26.34%的平均絕對百分比誤差(MAPE)。

The rise of fast fashion has overturned the business model of the traditional apparel industry, gradually moved towards a trend of short time to market, affordable prices and in line with fashion trends. The decline in product life cycles has led retailers to continuously introduce new products. Compared with other traditional industries, the apparel industry is a more consumer-oriented industry. However, the booming development of social media sites makes consumers vulnerable to change their preferences due to external factors. Therefore, accurate sales forecasting can avoid the risk of product shortages or sluggish sales. But for new products, the lack of historical sales data makes sales forecasting a tough problem.
This research proposes a two-stage method, including assisted design and intelligent forecasting approach. The first stage uses web crawler to browse a B2C apparel website, downloads the product names and sales quantities, disassembles the component elements of products from the product names, and reassembles them into new products to achieve assisted design. In the second stage, the clustering and classification methods are used to link the relationship between the combination of elements and the sales volume, and apply machine learning methods to predict the sales demand of new products. Finally, use Google Trends to collect the element composition of new products as external information indexes to adjust the forecasting value. The research results show that compared with other machine learning model (Random Forest and XGBoost), the intelligent forecasting approach can effectively reduce MSE by at least 45.79%, RMSE at least 26.35% and MAPE at least 26.34%.

Au, K. F., Choi T. M. & Yu Y. (2008). Fashion retail forecasting by evolutionary neural networks. International Journal of Production Economics, 114(2), 615-630.

Baghi, I., Gabrielli V. & Codeluppi V. (2013). Consumption practices of fast fashion products: A consumer-based approach. Journal of Fashion Marketing and Management, 17(2), 206-224.

Barnes, L. & Lea-Greenwood G. (2010). Fast fashion in the retail store environment. International Journal of Retail & Distribution Management, 38(10), 760-772.

Bhardwaj, V. & Fairhurst A. (2010). Fast fashion: response to changes in the fashion industry. The International Review of Retail, Distribution and Consumer Research, 20(1), 165-173.

Boone, T., Ganeshan R., Hicks R. L. & Sanders N. R. (2018). Can Google Trends Improve Your Sales Forecast? Production and Operations Management, 27(10), 1770-1774.

Breiman, L. (2001). Random Forests. Machine Learning, 45(1), 5-32.

Cachon, G. & Swinney R. (2011). The Value of Fast Fashion: Quick Response, Enhanced Design, and Strategic Consumer Behavior. Management Science, 57(4), 778-795.

Chen, T. & Guestrin C. (2016, August). XGBoost: A Scalable Tree Boosting System. Paper presented at the Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, California, USA.

Čiarnienė, R. & Vienažindienė M. (2014). Management of Contemporary Fashion Industry: Characteristics and Challenges. Procedia - Social and Behavioral Sciences, 156, 63-68.

Cui, R., Gallino S., Moreno A. & Zhang D. J. (2018). The Operational Value of Social Media Information. Production and Operations Management, 27(10), 1749-1769.

Ekambaram, V., Manglik K., Mukherjee S., Sajja S. S. K., Dwivedi S. & Raykar V. C. (2020, August). Attention based Multi-Modal New Product Sales Time-series Forecasting. Paper presented at the Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Virtual Event, California, USA.

Frank, C., Garg, A., Sztandera, L., & Raheja, A. (2003). Forecasting women's apparel sales using mathematical modeling. International Journal of Clothing Science and Technology, 15(2), 107-125.

Huang, H., & Liu, Q. (2017). Intelligent retail forecasting system for new clothing products considering stock-out. Fibres and Textiles in Eastern Europe, 25(1), 10-16.

Liaw, A., & Wiener, M. (2002). Classification and Regression by RandomForest. R News, 2(3), 18-22.

Lin, Y. T., Parlaktürk, A. K., & Swaminathan, J. M. (2014). Vertical Integration under Competition: Forward, Backward, or No Integration? Production and Operations Management, 23(1), 19-35.

Loureiro, A. L. D., Miguéis, V. L., & da Silva, L. F. M. (2018). Exploring the use of deep neural networks for sales forecasting in fashion retail. Decision Support Systems, 114, 81-93.

Mostard, J., Teunter, R., & De Koster, R. (2011). Forecasting demand for single-period products: A case study in the apparel industry. European Journal of Operational Research, 211(1), 139-147.

Ramos, P., & Oliveira, J. M. (2016). A Procedure for Identification of Appropriate State Space and ARIMA Models Based on Time-Series Cross-Validation. Algorithms, 9(4), 76.

Singh, P., Gupta, Y., Jha, N., & Rajan, A. (2019). Fashion Retail: Forecasting Demand for New Items. ArXiv, abs/1907.01960.

Sull, D., & Turconi, S. (2008). Fast Fashion Lessons. Business Strategy Review, 19(2), 4-11.

Sun, Z.-L., Choi, T.-M., Au, K.-F., & Yu, Y. (2008). Sales forecasting using extreme learning machine with applications in fashion retailing. Decision Support Systems, 46, 411-419.

Tehrani, A. F., & Ahrens, D. (2016). Improved Forecasting and Purchasing of Fashion Products based on the Use of Big Data Techniques. In R. Bogaschewsky, M. Eßig, R. Lasch, & W. Stölzle (Eds.), Supply Management Research: Aktuelle Forschungsergebnisse 2015 (pp. 293-312). Wiesbaden: Springer Fachmedien Wiesbaden.

Teucke, M., Ait-Alla, A., El-Berishy, N., Beheshti-Kashi, S., & Lütjen, M. (2016). Forecasting of Seasonal Apparel Products. In H. Kotzab, J. Pannek, & K.-D. Thoben (Eds.), Dynamics in Logistics (pp. 633-642). Cham: Springer International Publishing.

Thomassey, S., & Happiette, M. (2007). A neural clustering and classification system for sales forecasting of new apparel items. Applied Soft Computing, 7(4), 1177-1187.

Tin Kam, H. (1995, August). Random decision forests. Paper presented at the Proceedings of 3rd International Conference on Document Analysis and Recognition, Montreal, Quebec, Canada.

Tin Kam, H. (1998). The random subspace method for constructing decision forests. IEEE Transactions on Pattern Analysis and Machine Intelligence, 20(8), 832-844.

van Steenbergen, R. M., & Mes, M. R. K. (2020). Forecasting demand profiles of new products. Decision Support Systems, 139, [113401].

Whelan, C., Harrell, G., & Wang, J. (2015). Understanding the K-medians problem. Athens: The Steering Committee of The World Congress in Computer Science, Computer Engineering and Applied Computing (WorldComp).

Xia, M., & Wong, W. K. (2014). A seasonal discrete grey forecasting model for fashion retailing. Knowledge-Based Systems, 57, 119-126.

Yelland, P. M., & Dong, X. (2014). Forecasting Demand for Fashion Goods: A Hierarchical Bayesian Approach. In T.-M. Choi, C.-L. Hui, & Y. Yu (Eds.), Intelligent Fashion Forecasting Systems: Models and Applications (pp. 71-94). Berlin, Heidelberg: Springer Berlin Heidelberg

Yesil, E., Kaya, M., & Siradag, S. (2012, July). Fuzzy forecast combiner design for fast fashion demand forecasting. Paper presented at the 2012 International Symposium on Innovations in Intelligent Systems and Applications.

Yüzbaşıoğlu, O., & Küçükaydin, H. (2019). Forecasting with Ensemble Methods: An Application Using Fashion Retail Sales Data. Unpublished doctoral dissertation. MEF University Institute of Science and Technology, Istanbul, Turkey.