現今工業4.0與智慧製造(Smart Manufacturing)均是目前製造業發展的主要趨勢，本研究建立在透過AI的技術以非接觸式的方法建構一模擬SMT製程中Reflow生產過程的模型，並藉此預估各項產品的品質績效；本研究主要目的為透過特徵工程中特徵篩選的手法提升該預測模型的準確率，但在過程中卻發現工業變數難以定義彼此相關性的問題，因此本研究開發回焊爐特徵篩選方法的同時，也希望此方法能在往後相關研究中實際應用到工業變數的篩選中。
為解決工業變數無法定義彼此相關性並藉以找出重要特徵的問題，本研究使用了混合式特徵篩選技術(Hybrid Feature Selection)來取代傳統基於統計相關性指標的特徵篩選方法，並同時結合特徵篩選中過濾法(Filter Method)與包裝法(Wrapper Method)各自的優勢開發一基於增強式基因演算法混合式特徵篩選模型，為(Hybrid Genetic Algorithm – Bidirectional LSTM, HGA-Bi-LSTM)，該模型使用雙向長短期記憶神經網路進行基因演算法中配適度(Fitness)的評估。
根據數學模型中參數的設計，HGA-Bi-LSTM模型能在指定範圍內進行不同最佳特徵集合的搜索，同時也設計一輸入特徵篩選前的原特徵集合作為對照組，藉此來比較模型輸出之最佳特徵集合於預測模型上的績效是否優於未經任何篩選的特徵集合。在各種設定比較下，其最佳結果中演算法結果所輸出之模型準確率高達75.58%，相較對照組的60.08%，高了15.5%，並於Type I Error的判斷上也提升了22.39%，Type II Error雖未有明顯提升，但在檢驗產品個別品質規範中的表現卻是全部最佳特徵集合結果中最好的。因此可知本研究所開發之混合式特徵篩選模型能針對回焊爐溫度曲線中搜索出最佳特徵集合，並間接提升預測模型的整體效益。

Nowadays, Industry 4.0 and Smart Manufacturing are the future trends of the
current manufacturing industry. This research is based on a machine learning models
that can simulate the SMT reflow process, and the main purpose is to find a best feature
subset that can improve the prediction model’s accuracy. But industrial variables in this
research are hard to define each other's correlations, which means most traditional
correlation-based feature selection cannot be applied in this study, so the feature
selection approach proposed in this study have to solve the industrial variable problems
and finds the optimal feature subsets at the same time.
This research develops an HGA-Bi-LSTM method by using hybrid feature
selection technology, it can search different optimal feature subsets under various
parameter settings. This approach combines the advantanges of Filter method and
Wrapper method, first use the correlation-based feature selection (Filter Method) to
remove most redundant features, then search optimal feature subsets by genetic
algorithm (Wrapper Method), and use bidirectional LSTM to calculate the fitnees value
within the genetic algorithmns.
In the final comparison of the output results of various settings, the accuracy rate
of the prediction model in the best performing result is as high as 75.58%, which is
15.5% higher than the accuracy rate of 60.08% before using HGA-Bi-LSTM method.
Therefore, it can be seen that the hybrid feature screening model developed in this study
can search for the optimal feature subsets in the SMT reflow process, and improve the
overall benefit of the prediction model.

Althelaya, K. A., El-Alfy, E.-S. M., & Mohammed, S. (2018). Evaluation of bidirectional LSTM for short-and long-term stock market prediction. Paper presented at the 2018 9th international conference on information and communication systems (ICICS).
Amalu, E. H., Lau, W., Ekere, N., Bhatti, R., Mallik, S., Otiaba, K., & Takyi, G. (2011). A study of SnAgCu solder paste transfer efficiency and effects of optimal reflow profile on solder deposits. Microelectronic Engineering, 88(7), 1610-1617.
Amir, D. (1994). Expert system for SMT assembly. Paper presented at the Proceedings of the Surface Mount International Conference and Exposition-Technical Program.
Battiti, R. (1994). Using mutual information for selecting features in supervised neural net learning. IEEE Transactions on neural networks, 5(4), 537-550.
Biocca, P. (2013). Developing a reliable lead-free SMT assembly process. Kestar report.
Chang, Y.-M., Wei, C.-C., Chen, J., & Hsieh, P. (2019). An implementation of health prediction in SMT solder joint via machine learning. Paper presented at the 2019 ieee international conference on big data and smart computing (bigcomp).
Chuang, L.-Y., Chang, H.-W., Tu, C.-J., & Yang, C.-H. (2008). Improved binary PSO for feature selection using gene expression data. Computational Biology and Chemistry, 32(1), 29-38.
Chuang, L.-Y., Yang, C.-H., Wu, K.-C., & Yang, C.-H. (2011). A hybrid feature selection method for DNA microarray data. Computers in biology and medicine, 41(4), 228-237.
Chuang, L.-Y., Yang, C.-H., & Yang, C.-H. (2009). Tabu search and binary particle swarm optimization for feature selection using microarray data. Journal of computational biology, 16(12), 1689-1703.
Chuang, L.-Y., Yang, C.-S., Wu, K.-C., & Yang, C.-H. (2010). Correlation-based gene selection and classification using Taguchi-BPSO. Methods of information in medicine, 49(03), 254-268.
Coombs Jr, C. F. (2008). Printed circuits handbook: McGraw-Hill Education.
Corona, F., & Lendasse, A. (2005). Input selection and function approximation using the som: an application to spectrometric modeling. Paper presented at the WSOM'05, 5th Workshop on Self-Organizing Maps, Paris, France, September 5-8, 2005.
Doetsch, P., Kozielski, M., & Ney, H. (2014). Fast and robust training of recurrent neural networks for offline handwriting recognition. Paper presented at the 2014 14th international conference on frontiers in handwriting recognition.
Eiben, A. E., Raue, P.-E., & Ruttkay, Z. (1994). Genetic algorithms with multi-parent recombination. Paper presented at the International conference on parallel problem solving from nature.
Estévez, P. A., Perez, C. A., & Goles, E. (2003). Genetic input selection to a neural classifier for defect classification of radiata pine boards. Forest products journal, 53(7/8), 87-94.
Frohlich, H., Chapelle, O., & Scholkopf, B. (2003). Feature selection for support vector machines by means of genetic algorithm. Paper presented at the Proceedings. 15th IEEE International Conference on Tools with Artificial Intelligence.
Greff, K., Srivastava, R. K., Koutník, J., Steunebrink, B. R., & Schmidhuber, J. (2016). LSTM: A search space odyssey. IEEE transactions on neural networks and learning systems, 28(10), 2222-2232.
Hall, M. A. (1998). Correlation-based feature subset selection for machine learning. Thesis submitted in partial fulfillment of the requirements of the degree of Doctor of Philosophy at the University of Waikato.
Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780.
Hsu, H.-H., Hsieh, C.-W., & Lu, M.-D. (2011). Hybrid feature selection by combining filters and wrappers. Expert Systems with Applications, 38(7), 8144-8150.
Huang, Z., Yang, C., Zhou, X., & Huang, T. (2018). A hybrid feature selection method based on binary state transition algorithm and ReliefF. IEEE journal of biomedical and health informatics, 23(5), 1888-1898.
Hwang, J. S. (1996). Modern solder technology for competitive electronics manufacturing. McGraw-Hill, Inc, 1221 Avenue of the Americas, 3 rd Floor, New York, NY 10020, USA, 1996. 622.
Illés, B. (2010). Distribution of the heat transfer coefficient in convection reflow oven. Applied Thermal Engineering, 30(13), 1523-1530.
Jin, C., Jin, S.-W., & Qin, L.-N. (2012). Attribute selection method based on a hybrid BPNN and PSO algorithms. Applied Soft Computing, 12(8), 2147-2155.
Jing, S., Li, M., Li, X., & Yin, P. (2021). Optimization of reflow soldering temperature curve based on genetic algorithm. Energy Reports, 7, 772-782.
Kim, D., Koo, J., Kim, H., Kang, S., Lee, S. H., & Kang, J. T. (2019). Rapid fault cause identification in surface mount technology processes based on factory-wide data analysis. International Journal of Distributed Sensor Networks, 15(2), 1550147719832802.
Kumar, M., Husain, D., Upreti, N., & Gupta, D. (2010). Genetic algorithm: Review and application. Available at SSRN 3529843.
Lee, M.-C. (2009). Using support vector machine with a hybrid feature selection method to the stock trend prediction. Expert Systems with Applications, 36(8), 10896-10904.
Lee, N. C. (1999). Optimizing the reflow profile via defect mechanism analysis. Soldering & Surface Mount Technology.
Li, J., Cheng, K., Wang, S., Morstatter, F., Trevino, R. P., Tang, J., & Liu, H. (2017). Feature selection: A data perspective. ACM computing surveys (CSUR), 50(6), 1-45.
Li, Y., He, J., Won, D., & Yoon, S. W. (2021). Noncontact Reflow Oven Thermal Profile Prediction Based on Artificial Neural Network. IEEE Transactions on Components, Packaging and Manufacturing Technology, 11(12), 2229-2237.
Liu, H., & Yu, L. (2005). Toward integrating feature selection algorithms for classification and clustering. IEEE Transactions on knowledge and data engineering, 17(4), 491-502.
Liu, X., Krishnan, A., & Mondry, A. (2005). An entropy-based gene selection method for cancer classification using microarray data. BMC bioinformatics, 6(1), 1-14.
Liukkonen, M., Hiltunen, T., Havia, E., Leinonen, H., & Hiltunen, Y. (2009). Modeling of soldering quality by using artificial neural networks. IEEE Transactions on electronics packaging manufacturing, 32(2), 89-96.
Mitchell, M. (1998). An introduction to genetic algorithms: MIT press.
Mokler, S., Aspandiar, R., Byrd, K., Chen, O., Walwadkar, S., Tang, K. K., . . . Sane, S. (2016). The application of Bi-based solders for low temperature reflow to reduce cost while improving SMT yields in client computing systems. Paper presented at the Proceedings of SMTA International.
Muench, S., Bhat, D., Heindel, L., Hantschke, P., Roellig, M., & Kaestner, M. (2022). Performance Assessment of different Machine Learning Algorithm for Life-Time Prediction of Solder Joints based on Synthetic Data. Paper presented at the 2022 23rd International Conference on Thermal, Mechanical and Multi-Physics Simulation and Experiments in Microelectronics and Microsystems (EuroSimE).
Narendra, P. M., & Fukunaga, K. (1977). A branch and bound algorithm for feature subset selection. IEEE Transactions on computers, 26(09), 917-922.
Nasarian, E., Abdar, M., Fahami, M. A., Alizadehsani, R., Hussain, S., Basiri, M. E., . . . Acharya, U. R. (2020). Association between work-related features and coronary artery disease: A heterogeneous hybrid feature selection integrated with balancing approach. Pattern recognition letters, 133, 33-40.
Oh, I.-S., Lee, J.-S., & Moon, B.-R. (2004). Hybrid genetic algorithms for feature selection. IEEE Transactions on pattern analysis and machine intelligence, 26(11), 1424-1437.
Olah, C. (2015). Understanding lstm networks.
Oreski, S., & Oreski, G. (2014). Genetic algorithm-based heuristic for feature selection in credit risk assessment. Expert Systems with Applications, 41(4), 2052-2064.
Pan, J., Toleno, B. J., Chou, T. C., & Dee, W. J. (2006). The effect of reflow profile on SnPb and SnAgCu solder joint shear strength. Soldering & Surface Mount Technology.
Pudil, P., Novovičová, J., & Kittler, J. (1994). Floating search methods in feature selection. Pattern recognition letters, 15(11), 1119-1125.
Quinlan, J. R. (1986). Induction of decision trees. Machine learning, 1(1), 81-106.
Razali, N. M., & Geraghty, J. (2011). Genetic algorithm performance with different selection strategies in solving TSP. Paper presented at the Proceedings of the world congress on engineering.
Rehman, F., Asghar, R., Iqbal, K., Aman, A., & Nawaz, A. A. (2020). An experimental study of carbonyl powder power inductor cracking during reflow process. Soldering & Surface Mount Technology.
Saidi, R., Bouaguel, W., & Essoussi, N. (2019). Hybrid feature selection method based on the genetic algorithm and pearson correlation coefficient. In Machine learning paradigms: theory and application (pp. 3-24): Springer.
Sarvar, F., & Conway, P. P. (1998). Effective modeling of the reflow soldering process: basis, construction, and operation of a process model. IEEE Transactions on Components, Packaging, and Manufacturing Technology: Part C, 21(2), 126-133.
Schuster, M., & Paliwal, K. K. (1997). Bidirectional recurrent neural networks. IEEE transactions on Signal Processing, 45(11), 2673-2681.
Siedlecki, W., & Sklansky, J. (1993). A note on genetic algorithms for large-scale feature selection. In Handbook of pattern recognition and computer vision (pp. 88-107): World Scientific.
Straubinger, D., Bozsóki, I., Illés, B., & Géczy, A. (2019). Analytical solution of heat distribution inside a printed circuit board during vapour phase soldering. Paper presented at the 2019 42nd International Spring Seminar on Electronics Technology (ISSE).
Straubinger, D., Bozsóki, I., Illes, B., Krammer, O., Bušek, D., & Geczy, A. (2020). Heat transfer aspects of condensation during vapour phase soldering on aligned PCB-based surfaces. Soldering & Surface Mount Technology, 32(4), 247-252.
Su, Y., Srihari, K., & Emerson, C. (1997). A profile identification system for surface mount printed circuit board assembly. Computers & industrial engineering, 33(1-2), 377-380.
Tan, F., Fu, X., Zhang, Y., & Bourgeois, A. G. (2008). A genetic algorithm-based method for feature subset selection. Soft Computing, 12(2), 111-120.
Terng, E. F., Yeoh, S. C., Tong, K. C., & Yeo, K. S. (2020). Data analysis on SMT reflow oven with SECS/GEM communication protocol. Paper presented at the 2020 IEEE 10th Symposium on Computer Applications & Industrial Electronics (ISCAIE).
Tsoukalas, L. H., & Uhrig, R. E. (1996). Fuzzy and neural approaches in engineering: John Wiley & Sons, Inc.
Wei, G., Zhao, J., Feng, Y., He, A., & Yu, J. (2020). A novel hybrid feature selection method based on dynamic feature importance. Applied Soft Computing, 93, 106337.
Zaremba, W., Sutskever, I., & Vinyals, O. (2014). Recurrent neural network regularization. arXiv preprint arXiv:1409.2329.
Zhang, H., & Sun, G. (2002). Feature selection using tabu search method. Pattern recognition, 35(3), 701-711.
許哲昇. (2017). 使用長短期記憶深度學習 之機器剩餘可用壽命預估. (碩士). 國立中央大學, 桃園縣. Retrieved from https://hdl.handle.net/11296/7er8xs
蔡聰男. (2003). 自適應式表面黏著製程品質預測控制系統之發展. (博士). 國立成功大學, 台南市. Retrieved from https://hdl.handle.net/11296/84td46