隨著近年來電腦軟硬體、網路的快速發展，儲存裝置之普及化及大容量化使得計算科學在各個領域上所佔之地位水漲船高，且利用電腦運算、工程實驗或網路蒐集到之資料量亦都大幅度的增加且能保存，保存下來的資料量龐大到傳統的資料處理方式已無法處理，需要新型態的資料處理與挖掘之方式，因此資料探勘的發展十分迅速。
本文主要目的為開發一資料探勘之標準流程，配合Python軟體之強大擴充性及平易近人之直譯語言，完成使用彈性倒傳遞類神經網路演算法，針對大量輸入資料建立多輸出目標之預測模型，以及配合基因演算法以求解所建立之預測模型之最佳參數組合解。利用收集自參考文獻之130組資料，每組資料內含4個輸入參數與2個輸出目標，以所建立相關之預測模型進行預測，結果顯示兩個輸出目標預測值之均方根相對誤差分別為2.27%以及3.97%，驗證了本文所建立之預測模型之可靠度；同時利用Rastrigin function驗證本資料探勘平台於搜尋最佳參數組合部分之可執行性與正確性，也對所使用之兩種演算法內，重要參數變異對結果之影響做逐一討論。
最後針對本實驗室所模擬之64筆FCBGA封裝翹曲之數據，從類神經演算法到基因演算法，進行了整組最佳化流程。從建立預測模型時，三輸出變量之均方根相對誤差分別為1.72%、2.42%及4.51%，確立了模型之可靠度。再配合基因演算法進行多目標之最小值最佳化，所得之結果分別為483.28、359.33及176.39，三輸出值也皆低於已知數據中之最小值，再次驗證了建立之最佳化流程之可靠度，以期本法能提供給相關業界推廣及使用。

In recent years, due to the rapid development of computational science, software, and the internet, the popularity of storage device and high-capacity make the computational science rising greatly in various fields. Using computer to calculate, engineering experiment or information collecting from internet have increased greatly and conserved. The amount of information is so huge that the traditional data processing can’t be handled. It needs a new type of data processing, so the development of data mining is very rapid.
The purpose of this study is to develop a standard operating procedure of data mining, written by Python, which is highly expandable and easy to be compiled. Such that the predicting model corresponding to a massive amount of input engineering data can be obtained by using artificial neural network algorithms and optimized the engineering problem by using genetic algorithms. In this study, 130 sets of data collected from an existed literature, each set of data contains four input parameters and two output targets, were inputted into the proposed data mining procedure to establish the relevant predicting model. The predicted results showed that the mean square root of relative error of the two outputs target is 2.27% and 3.97%, respectively, which verifies the reliability of the predicting model established by the proposed data mining procedure. Furthermore, this study also uses Rastrigin function to verify the feasibility and correctness of the proposed data mining procedure in searching for the optimal combination of input parameters. This study also discusses the variations of the important control parameters of the two algorithms used in the proposed data mining procedure on the impact of the results.
Finally, this study established a predicting model and optimizing the multi-objective target, using 64 sets of data simulated by ANSYS. The predicted results showed that the RMSE of the three outputs target are 1.72%、2.42% and 4.51%, and verifies the reliability of the predicting model. By using genetic algorithms, we also got the results of multi-objective optimization which are 483.28、359.33 and 176.39, all of those are lower than the minimum of the datasets we got. Verifying the multi-objective optimization we established again.

論文審定書 i
誌 謝 ii
摘 要 iii
Abstract iv
目錄 v
表目錄 ix
圖目錄 xi
第一章 1
1.1 研究背景 1
1.2 資料探勘簡介 2
1.3 文獻回顧 3
1.3.1 類神經網路演算法演進歷程 3
1.3.2 類神經網路演算法於工程上之應用 4
1.3.3 基因演算法演進歷程與其應用 5
1.4 研究動機與目的 6
1.5 全文架構 7
第二章 9
2.1 類神經網路演算法原理簡介 9
2.1.1多層前饋式類神經網路 11
2.1.2倒傳遞類神經網路演算法 12
2.2 基因演算法原理簡介 13
2.2.1編碼 14
2.2.2初始母代族群 14
2.2.3適應值 14
2.2.4選擇與複製 15
2.2.5交配 15
2.2.6突變 16
2.2.7菁英策略 16
2.3 Python語言軟體介紹 17
第三章 26
3.1 資料探勘演算法整體建立與運作流程 26
3.2 資料輸入與資料預處理 26
3.2.1數據輸入 26
3.2.2資料預處理 27
3.3 類神經網路預測模型建立 27
3.3.1 類神經網路演算法運作流程 28
3.3.2 類神經網路演算法參數設定 28
3.4 基因演算法找最佳參數組合 30
3.4.1 基因演算法運作流程 30
3.4.2 基因演算法參數設定 31
3.5演算法運算結果的呈現與輸出 32
3.5.1圖像化顯示 32
3.5.2 輸出存檔格式 33
第四章 44
4.1 驗證演算法可靠度原始數據 44
4.2 驗證演算法可靠度預測模型結果 46
4.3 類神經網路演算法之參數對預測模型影響 47
4.3.1 隱藏層與神經元 47
4.3.2 Dropout層 49
4.3.3 初始輸入資料比例 49
4.4 驗證基因演算法之最佳參數組合結果 49
4.4.1 驗證基因演算法 49
4.4.2 基因演算法之族群個體數量對結果之影響 50
4.4.3 基因演算法之交配與突變機率對結果之影響 51
4.4.4 基因演算法之最大迭代次數對結果之影響 51
4.5 對FCBGA數據之資料探勘 51
4.5.1類神經網路建模 51
4.5.2基因演算法之多變量最佳化 52
第五章 78
5.1 類神經網路預測值與反應曲面法預測值之結果差異分析 78
5.2 類神經網路中參數對結果之差異分析 79
5.2.1 隱藏層與神經元 79
5.2.2 Dropout層 80
5.2.3 初始輸入資料比例 80
5.3 基因演算法及其參數之結果差異分析 80
5.3.1 族群群體個數 80
5.3.2 交配與突變之機率 81
5.3.3 最大迭代次數 82
5.4 FCBGA數據之類神經與基因演算法 82
第六章 86
6.1 結論 86
6.2 未來展望 88
參考文獻 90
附錄一 93
附錄二 97

[1] E. W. T. Ngai, L. Xiu, and D. C. K. Chau, "Application of data mining techniques in customer relationship management: A literature review and classification," Expert Systems with Applications, vol. 36, no. 2, Part 2, pp. 2592-2602, 2009.
[2] U. M. Fayyad, G. Piatetsky-Shapiro, and P. Smyth, "From data mining to knowledge discovery: an overview," in Advances in knowledge discovery and data mining, M. F. Usama, P.-S. Gregory, S. Padhraic, and U. Ramasamy, Eds.: American Association for Artificial Intelligence, 1996, pp. 1-34.
[3] W. S. McCulloch and W. Pitts, "A logical calculus of the ideas immanent in nervous activity," The bulletin of mathematical biophysics, journal article vol. 5, no. 4, pp. 115-133, 1943.
[4] D. O. Hebb, Organization of Behavior. New York: Wiley & Sons, 1949.
[5] B. Farley and W. Clark, "Simulation of self-organizing systems by digital computer," Transactions of the IRE Professional Group on Information Theory, vol. 4, no. 4, pp. 76-84, 1954.
[6] F. Rosenblatt, "The perceptron: A probabilistic model for information storage and organization in the brain," Psychological Review, vol. 65, no. 6, pp. 386-408, 1958.
[7] S. U. S. E. Laboratories, B. Widrow, U. S. O. o. N. Research, U. S. A. S. Corps, U. S. A. Force, and U. S. Navy, Adaptive "adaline" neuron using chemical "memistors.". 1960.
[8] M. L. Minsky and S. Papert, "Perceptrons," ed, 1969.
[9] P. Werbos, "Beyond Regression: New Tools for Prediction and Analysis in the Behavioral Sciences," Harvard University, 1974.
[10] J. J. Hopfield, "Neural networks and physical systems with emergent collective computational abilities," Proceedings of the National Academy of Sciences, vol. 79, no. 8, pp. 2554-2558, 1982.
[11] D. E. Rumelhart, G. E. Hinton, and R. J. Williams, "Parallel Distributed Processing: Explorations in the Microstructure of Cognition," vol. 1, ed, 1986.
[12] D. E. Rumelhart, G. E. Hinton, and R. J. Williams, "Learning representations by back-propagating errors," Nature, 10.1038/323533a0 vol. 323, no. 6088, pp. 533-536, 1986.
[13] M. Riedmiller and H. Braun, "RPROP - A Fast Adaptive Learning Algorithm," Proc. of ISCIS VII), 1992.
[14] D. C. Park, M. A. El-Sharkawi, R. J. Marks, L. E. Atlas, and M. J. Damborg, "Electric load forecasting using an artificial neural network," IEEE Transactions on Power Systems, vol. 6, no. 2, pp. 442-449, 1991.
[15] 林文賜, 周天穎, and 林昭遠, "「科技短文」應用監督性類神經網路於衛星影像分類技術之探討," (in 繁體中文), 航測及遙測學刊, vol. 6, no. 1, pp. 41-58, 2001.
[16] 張憲國, 錢維安, and 何良勝, "應用類神經網路在台灣東岸海域颱風波浪推算之研究," (in 繁體中文), 海洋工程學刊, vol. 3, no. 1, pp. 73-96, 2003.
[17] P. C. Lai, T. C. Chiu, and G. S. Shen, "Design optimization for wafer level package reliability by using artificial neural network," in 2013 8th International Microsystems, Packaging, Assembly and Circuits Technology Conference (IMPACT), , pp. 172-175. 2013
[18] 賴柏辰, "應用類神經網路及遺傳演算法於晶圓級封裝之結構最佳化," 機械工程學系, 成功大學, 2013.
[19] P. Lall, S. Deshpande, and L. Nguyen, "ANN based RUL assessment for copper-aluminum wirebonds subjected to harsh environments," in 2016 IEEE International Conference on Prognostics and Health Management (ICPHM), pp. 1-10, 2016.
[20] Z. Lang, W. Hai, and S. X. D. Tan, "Fast stress analysis for runtime reliability enhancement of 3D IC using artificial neural network," in 2016 17th International Symposium on Quality Electronic Design (ISQED), pp. 173-178, 2016.
[21] C. Darwin, On the Origin of Species by Means of Natural Selection, Or, The Preservation of Favoured Races in the Struggle for Life. J. Murray, 1859.
[22] G. Mendel, "Versuche über Pflanzen-Hybriden," Verhandlungen des naturforschenden Vereines in Brünn., vol. Bd.4 (1865-1866), pp. 3-47, 1865.
[23] J. H. Holland, Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control and Artificial Intelligence. MIT Press, p. 228, 1992.
[24] D. E. Goldberg and J. H. Holland, "Genetic Algorithms and Machine Learning," Machine Learning, journal article vol. 3, no. 2, pp. 95-99, 1988.
[25] J. Horn, N. Nafpliotis, and D. E. Goldberg, "A niched Pareto genetic algorithm for multiobjective optimization," in Proceedings of the First IEEE Conference on Evolutionary Computation. IEEE World Congress on Computational Intelligence, pp. 82-87 vol.1, 1994.
[26] N. Srinivas and K. Deb, "Muiltiobjective Optimization Using Nondominated Sorting in Genetic Algorithms," Evolutionary Computation, vol. 2, no. 3, pp. 221-248, 1994.
[27] E. Zitzler and L. Thiele, "Multiobjective evolutionary algorithms: a comparative case study and the strength Pareto approach," IEEE Transactions on Evolutionary Computation, vol. 3, no. 4, pp. 257-271, 1999.
[28] J. D. Schaffer, "Some experiments in machine learning using vector evaluated genetic algorithms (artificial intelligence, optimization, adaptation, pattern recognition)," Vanderbilt University, 1984.
[29] P. Hajela and C.-Y. Lin, "Genetic search strategies in multicriterion optimal design," Structural optimization, journal article vol. 4, no. 2, pp. 99-107, 1992.
[30] 李聰穎, 林義傑, 詹榮茂, and 陳俊隆, "基因演算法應用於時間電價用戶契約容量選定," ed: 明新學報, 2005.
[31] 葉進儀, 林彣珊, and 朱慶餘, "應用平行基因演算法改善護理人員排班品質," (in 繁體中文), 品質學報, vol. 14, no. 3, pp. 337-350, 2007.
[32] M. Qiu, Z. Ming, J. Li, K. Gai, and Z. Zong, "Phase-Change Memory Optimization for Green Cloud with Genetic Algorithm," IEEE Transactions on Computers, vol. 64, no. 12, pp. 3528-3540, 2015.
[33] 李偉齊, "光纖雙向收發元件之材料熱傳導係數變異對散熱效果之影響," vol. 台灣, ed. 機械與機電工程學系: 國立中山大學, 2016.
[34] F. Günther and S. Fritsch, "neuralnet Training of Neural Networks," The R Journal, vol. 2, no. 1, pp. 30-38, 2010.
[35] L. Zhang, G. Subbarayan, B. C. Hunter, and D. Rose, "Response surface models for efficient, modular estimation of solder joint reliability in area array packages," Microelectronics Reliability, vol. 45, no. 3, pp. 623-635, 2005/03/01/ 2005.
[36] J. Feldman and R. Rojas, Neural Networks: A Systematic Introduction. Springer Berlin Heidelberg, 1996.
[37] 徐永吉. 林昇甫, 遺傳演算法及其應用 五南圖書出版公司, 2009.
[38] 袁克倫、楊孟穎, It's Django：用Python迅速打造Web應用. 臺北市: 碁峰資訊, 2015.
[39] K. Hornik, M. Stinchcombe, and H. White, "Multilayer feedforward networks are universal approximators," Neural Netw., vol. 2, no. 5, pp. 359-366, 1989.
[40] 類神經網路模式應用與實作. 儒林, 2009.
[41] J. Heaton, Introduction to Neural Networks for Java, 2nd Edition. Heaton Research, Inc., p. 440, 2008.
[42] D. R. Hush and B. G. Horne, "Progress in supervised neural networks," IEEE Signal Processing Magazine, vol. 10, no. 1, pp. 8-39, 1993.