隨著數位時代的來臨，加上在全球化市場競爭之影響下，現今消費者需求已從過去大量標準化轉變為高度客製化，這也迫使各產業亟需尋求產業經營策略的轉型，以應對當代需求多變的產業環境。正因為如此，智慧製造與智慧生產促成近幾年廣受討論的工業4.0之核心議題。以智慧製造作為發展主軸，結合物聯網(Internet of things, IOT)、網宇實體系統(Cyber physical system, CPS)、人工智慧(Artifical intelligence, AI)及大數據(Big data, BD)等技術來提升生產及管理效率，並輔助管理者做出應對的決策。
本研究以一家板金加工(Sheet metal fabrication)公司為案例分析對象，利用智慧機上盒(Smart manufacturing box, SMB)所蒐集的資料為基礎，建構迴歸決策樹(Regression tree, RT)、隨機森林(Random forest, RF)及支援向量迴歸(Support vector regression, SVR)三種迴歸預測模型，藉以解決過去板金加工業工時難以估計之問題。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(Grid search)結合交叉驗證(Cross validation)之方法找出各模型之最佳參數組合。最後比較三種迴歸預測模型於各製程之表現，並選擇各製程之最佳預測模型。
實驗結果表明，雷射製程最佳預測模型為支援向量迴歸，其餘三個製程(沖床、折床及剪床)則是隨機森林最好；雷射製程平均絕對百分比誤差約為14%，其餘三個製程平均誤差約為11%。該模式相較於原始平均約34%的人工估計誤差，對於接近100%客製化的板金加工產業，已是足夠輔助管理人員判斷工時的一個標準依據，也達成本研究欲建構可靠的製程加工時間預測模型之目的。

With the advent of the digital age, and under the influence of global market competition, today's consumer demand has changed from a large number of standardization in the past to a high degree of customization, which also forces the industries to seek the transformation of industrial management strategies to cope with the changing industrial environment. Because of this, intelligent manufacturing and intelligent production have become the core issue of industry 4.0, which has been widely discussed in recent years. With smart manufacturing as the development axis, combined with Internet of things (IoT), Cyber physical system (CPS), Artificial intelligence (AI) and Big data (BD) techniques to improve production and management efficiency, and assist managers to make corresponding decisions.
This study takes a sheet metal fabrication company as the case study object. Based on the data collected from the Smart machine box (SMB), three regression prediction models, namely Regression tree (RT), Random forest (RF) and Support vector regression (SVR), are constructed to solve the problem of hard estimation of sheet metal processing time in the past. Firstly, the collected data are preprocessed, and then the best combination of parameters of each model is found by Grid search combined with cross validation. Finally, the performance of the three regression models in each process was compared, and the best prediction model for each process was selected.
The experimental results show that the best prediction model for the laser process is SVR, and the other three processes (pressing, folding, and cutting) are RF. For instance, the maen absolute percentage error of the laser process is about 14%, and the other three processes average error is about 11%. Compared with the original average of about 34% of the artificial estimation error, this model is a standard basis for assisting managers to judge the working hours for the nearly 100% customized sheet metal fabrication industry, and it also achieves the purpose of this research to construct reliable processing time prediction models based on BD analytics.

經濟部工業局，2019，智慧機上盒(SMB)輔導計畫，available at http://www.tami.org.tw/wisdom_machine/wisdom_machine-735-1.pdf(取得日期：2021.05.02)

經濟部統計處，2020，工業產銷存動態調查，available at https://dmz26.moea.gov.tw/GMWeb/investigate/InvestigateDB.aspx(取得日期：2021.05.01)

楊大和、李兆華，2021，精實管理4.0之自働生產管理環，工業工程雜誌，2021年5月，available at https://sites.google.com/ciie.org.tw/column501/精實管理4-0之自働生產管理環?authuser=2 (取得日期：2021.06.12)

簡禎富，2019，台灣企業邁向智慧製造與數位決策的戰略，台北：天下雜誌股份有
限公司。

簡禎富、許嘉裕，2014，資料挖礦與大數據分析。台北：前程文化。

黎安松，2014，實用板金學。新北：全華圖書。

李晟源，2015，支援向量迴歸利用行政區域分類預測房屋價格，台北市立大學資訊科學研究所學位論文，頁1-44。

林焜詳，2016，支撐向量機與隨機森林應用於颱風時雨量預報之比較，國立成功大學水利及海洋工程研究所學位論文，頁1-105。

許宸綺，2020，用心跳間隔時間之自身相關性質診斷充血性心力衰竭-區辨分析、決策樹與隨機森林之比較，東海大學統計系研究所學位論文，頁1-36。

Berry, M. J., & Linoff, G. S. (2004). Data Mining Techniques: For Marketing, Sales, and Customer Relationship Management. John Wiley & Sons, New York.

Breiman, L., Friedman, J., Stone, C. J., & Olshen, R. A. (1984). Classification and Regression Trees. Chapman & Hall, New York.

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

Coussement, K., & Van den Poel, D. (2008). Churn prediction in subscription services: An application of support vector machines while comparing two parameter-selection techniques. Expert Systems with Applications, 34(1), 313-327.

Drucker, H., Burges, C. J., Kaufman, L., Smola, A., & Vapnik, V. (1997). 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.

Gerhardt, B., Griffin, K., & Klemann, R. (2012). Unlocking value in the fragmented
world of big data analytics. Cisco Internet Business Solutions Group, 7.

Gunn, S. R. (1998). Support vector machines for classification and regression. ISIS
Technical Report, 14(1), 5-16.

Guo, L., Chehata, N., Mallet, C., & Boukir, S. (2011). Relevance of airborne lidar and
multispectral image data for urban scene classification using Random Forests. ISPRS Journal of Photogrammetry and Remote Sensing, 66(1), 56-66.

Hastie, T., Tibshirani, R., & Friedman, J. (2009). The Elements of Statistical Learning:
Data Mining, Inference, and Prediction. Springer, New York.

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

Hsu, C. W., Chang, C. C., & Lin, C. J. (2003). A practical guide to support vector
classification. Tech. Rep., Department of computer science, National Taiwan University.

Lewis, C. D. (1982). Industrial and Business Forecasting Methods: A Practical Guide to Exponential Smoothing and Curve Fitting. Butterworth Scientifi, London.

Lewis, R. J. (2000, May). An introduction to classification and regression tree (CART)
analysis. In Annual Meeting of the Society for Academic Emergency Medicine in San Francisco, California (Vol. 14).

McAfee, A., Brynjolfsson, E., Davenport, T. H., Patil, D. J., & Barton, D. (2012). Big
data: the management revolution. Harvard Business Review, 90(10), 60-68.

McGill, R., Tukey, J. W., & Larsen, W. A. (1978). Variations of box plots. The American
Statistician, 32(1), 12-16.

Prasad, A. M., Iverson, L. R., & Liaw, A. (2006). Newer classification and regression
tree techniques: bagging and random forests for ecological prediction. Ecosystems, 9(2), 181-199.

Ray, S. (2019, February). A quick review of machine learning algorithms. In 2019
International Conference on Machine Learning, Big Data, Cloud and Parallel Computing (COMITCon) (pp. 35-39). IEEE.

Sagiroglu, S., & Sinanc, D. (2013, May). Big data: A review. In 2013 International
Conference on Collaboration Technologies and Systems (CTS) (pp. 42-47). IEEE.
Song, Y. Y., & Ying, L. U. (2015). Decision tree methods: applications for classification and prediction. Shanghai Archives of Psychiatry, 27(2), 130.

Tsai, T. L., & Lee, C. Y. (2017). Data mining for yield improvement of photo spacer
process in color filter manufacturing. Procedia Manufacturing, 11, 1958-1965.

Vincenzi, S., Zucchetta, M., Franzoi, P., Pellizzato, M., Pranovi, F., De Leo, G. A., &
Torricelli, P. (2011). Application of a Random Forest algorithm to predict spatial distribution of the potential yield of Ruditapes philippinarum in the Venice lagoon, Italy. Ecological Modelling, 222(8), 1471-1478.

Yu, X., Qi, Z., & Zhao, Y. (2013). Support vector regression for newspaper/magazine
sales forecasting. Procedia Computer Science, 17, 1055-1062.