臺灣博碩士論文加值系統

採購交期管理係指採購人員在下單給供應商後，針對採購品項的交貨期進行跟催及管理。若供應商交期延誤頻率高，則可能會造成公司競爭力下降、產品交貨延遲及客戶滿意度下降的問題。本研究使用採購相關資料做為原始資料集，並與採購人員進行供應商交期延誤的定義討論，再對資料集做前處理及特徵擴充。透過資料探勘分類演算法，本研究使用單純貝氏分類（Naïve Bayes）、隨機森林（Random Forests）、XGBoost（eXtreme Gradient Boosting）、K-近鄰演算法（K Nearest Neighbor Classfication）、Light GBM（Light Gradient Boosting Machine）進行供應商交貨延遲預測模型建立及分析比較，找出較佳的分類預測方法。
造成供應商交期延誤的原因，可能在於人員、料件、供應商本身，或是數種特徵的結合體，本研究使用關聯規則（Association Rule）進行供應商交期延誤的關鍵因子探討，並將結果提供給採購人員進行改善的依據。

Supplier Delivery Management refers to the procurement and management of the procurement items after placing orders with the supplier. If the supplier's delivery delay is high, it may cause the company's competitiveness to decline, delayed product delivery and decreased customer satisfaction.
This study uses procurement-related data as the original dataset, and discusses the definition of supplier delivery delay with the purchasing staff, and then pre-processes and expands the features of the dataset. Through the data mining classification algorithm, this study uses Naïve Bayes, Random Forests, XGBoost, KNN, LGBM to establish and analyze the model to predict the supplier's delivery delay.
The reasons for supplier delivery delays may be due to personnel, materials, suppliers themselves, or a combination of several characteristics. This study also uses the Association Rule to explore key factors for supplier delivery delays, and provides the results to the purchasing staff as a basis for improvement

摘要 i
Abstract ii
致謝 iii
目錄 iv
圖目錄 vi
表目錄 viii
一、緒論 1
1.1研究背景 1
1.2研究動機與目的 1
1.3研究步驟 2
1.4論文架構 4
二、文獻探討 6
2.1採購管理 6
2.2交期管理 7
2.3資料探勘技術 7
2.4關聯規則 8
2.5資料分類方法 9
2.5.1單純貝氏分類模型（Naïve Bayes） 10
2.5.2隨機森林（Random Forest） 10
2.5.3 XGBoost（eXtreme Gradient Boosting） 11
2.5.4 K-近鄰演算法（K Nearest Neighbor Classification） 11
2.5.5 Light GBM（Light Gradient Boosting Machine） 12
2.6相關研究 12
2.6.1以資料探勘分析多樣少量訂貨生產採購流程時間之研究 12
2.6.2資料探勘技術在晶圓針測誤宰分析之應用 13
三、研究方法與架構 14
3.1研究架構 14
3.2研究對象現況說明 14
3.3資料集 16
3.3.1資料前處理 17
3.4資料探勘工具 26
3.4.1 Anaconda 3 26
3.4.2 Python語言 26
3.4.3 Scikit-Learn 27
四、實驗與結果分析 29
4.1實驗一：供應商交期延遲關鍵因子分析 29
4.1.1問題定義 29
4.1.2關聯規則函式庫 30
4.1.3資料準備及前處理 30
4.1.4關鍵因子資料探勘 33
4.2實驗二：建立預測供應商交期是否延誤分類模型 40
4.2.1問題定義 40
4.2.2資料集與特徵 40
4.2.3模型評估方法 40
4.2.4建立分類預測模型 42
五、結論及未來展望 49
5.1 結論 49
5.2 未來展望 50
參考文獻 51

[1] R. Agrawal, T. Imieliński, and A. Swami, Mining association rules between sets of items in large databases. ACM SIGMOD Record, 1993. 22(2): pp. 207-216.
[2] R. Agrawal and R. Srikant, Fast Algorithms for Mining Association Rules in Large Databases, in Proceedings of the 20th International Conference on Very Large Data Bases. 1994, Morgan Kaufmann Publishers Inc. pp. 487-499.
[3] N.S. Altman, An Introduction to Kernel and Nearest-Neighbor Nonparametric Regression. The American Statistician, 1992. 46(3): pp. 175-185.
[4] L. Breiman, Bagging Predictors. Machine Learning, 1996. 24(2): pp. 123-140.
[5] L. Breiman, Random Forests. Machine Learning, 2001. 45(1): pp. 5-32.
[6] L. Breiman, et al., Classification and Regression Trees. 1984: Taylor & Francis.
[7] T. Chen and C. Guestrin, XGBoost: A Scalable Tree Boosting System. 2016.
[8] J. Davis and M. Goadrich, The Relationship Between Precision-Recall and ROC Curves. 2006, ACM Press.
[9] U. Fayyad, G. Piatetsky-Shapiro, and P. Smyth, The KDD process for extracting useful knowledge from volumes of data. Communications of the ACM, 1996. 39(11): pp. 27-34.
[10] W.J. Frawley, G. Piatetsky-Shapiro, and C.J. Matheus, Knowledge discovery in databases: an overview. AI Magazine, 1992. 13(3): pp. 57.
[11] J.H. Friedman, Greedy function approximation: A gradient boosting machine. Ann. Statist., 2001. 29(5): pp. 1189-1232.
[12] G. Ke, et al., LightGBM: a highly efficient gradient boosting decision tree, in Proceedings of the 31st International Conference on Neural Information Processing Systems. 2017, Curran Associates Inc.: Long Beach, California, USA. pp. 3149-3157.
[13] S.-L. Org. Scikit-Learn algorithm cheat-sheet. Available from: https://scikit-learn.org/stable/index.html.
[14] D. Powers and Ailab, Evaluation: From precision, recall and F-measure to ROC, informedness, markedness & correlation. J. Mach. Learn. Technol, 2011. 2: pp. 2229-3981.
[15] I. Rish, An empirical study of the naive Bayes classifier, in IJCAI 2001 workshop on empirical methods in artificial intelligence. 2001. pp. 41-46.
[16] P.-N. Tan, Introduction to data mining, M. Steinbach and V. Kumar, Editors. 2006, Boston : Pearson Addison Wesley: Boston.
[17] B. Thuraisingham, A primer for understanding and applying data mining. IT Professional, 2000. 2(1): pp. 28-31.
[18] Wikipedia. Anaconda. Available from: https://zh.wikipedia.org/wiki/Anaconda_(Python发行版).
[19] Wikipedia. Python. Available from: https://zh.wikipedia.org/wiki/Python.
[20] Wikipedia. scikit-learn. Available from: https://en.wikipedia.org/wiki/Scikit-learn.
[21] 丁世杰, 資料探勘技術在晶圓針測誤宰分析之應用, in 理學院應用科技學程. 2011, 國立交通大學. pp. 48.
[22] 王忠宗, 採購管理實務, 空大, 2006.
[23] 肖瀟, 賺錢公司的採購學, 大是文化, 2018.
[24] 林大貴, Python+ Spark 2.0+Hadoop機器學習與大數據分析實戰, 博碩文化, 2016.
[25] 林秋堂, 採購概論與實務 = An Introduction to Purchasing and Practice, 新文京, 2003.
[26] 財經. 交期管理：如何確保供應商如期交貨？. [Internet] 2015; Available from: https://kknews.cc/finance/gqqmvpy.html.
[27] 曾素真, 以資料探勘分析多樣少量訂貨生產採購流程時間之研究, in 工業與資訊管理學系專班. 2007, 國立成功大學: 台南市. pp. 67.
[28] 黃錦江, 以ISO-9001精神建立交期管理系統之研究---微機電公司為例, in 管理學院工業工程與管理學程. 2006, 國立交通大學. pp. 129.
[29] 嶋津司, 採購管理, 譯. 簡錦川, 書泉, 2001.