發光二極體(Light Emitting Diode)取代原有照明設備的基本功能已不能滿足人類對發光二極體的期望；發光二極體產品因人類對終端電子產品的需求，於於2012年至2016年有爆炸性的需求量，因此圖案化藍寶石基板(Pattern Sapphire Substrate)製程導入，也改變了發光二極體的原有製程。現有圖案化藍寶石基板生產工廠如何依客戶不同的規格需求，準確、快速的調整製程參數，讓測試的成本與時間降至最低，便能有效地降低生產成本與提升企業的毛利。但在圖案化藍寶石基板製程，加上製程固有的變異，傳統的調整方式是透過製程工程師個人的經驗進行調整，且不同製程工程師的從業年資及經驗，往往會有參數調整次數的不同，若調整方向錯誤，則造成不必要的成本的損失。本研究利用類神經網路(Artificial Neural Network)、支援向量機(Support Vector Machine)及複迴歸分析(Multiple Regression Analysis)發展一套圖案化藍寶石基板製程結果的預測模型，並以某圖案化藍寶石基板代工廠之生產資料比較個模型預測結果之差異。經由本研究實證結果，於圖案化藍寶石基板製程結果的預測模型，類神經網路、支援向量機、複迴歸分析，皆有良好的預測效果。因此若有製程調整需求或是新產品開發，製程工程師皆可透過預測模型的結果，做為實際參數調整的參考方案。

The invention of Light Emitting Diode (LED) to replace the original lighting equipment is the original goal of human beings. However, from 2012 to 2016, electronic terminal products have become the most important demand of the LED industry. The update rate of electronic terminal products is once a year, so cost and time become the key.
Patterned Sapphire Substrate (PSS) process can directly increase brightness by 10% to 25% and has changed the original LED manufacturing process. PSS helps accurately and quickly adjust the process parameters according to the specifications of the final products. This allows for the reduction of the testing time and the production cost. However, there is an inherent trial and error phenomenon in the adjustment method. Even though, the traditional adjustment is supplemented by the personal experience of the process engineer, years of experience is required for accurate production results. And different process engineers often have different suggestions for the parameter adjustments.
This thesis implements three predictive models for the PSS results using Artificial Neural Network (ANN), Support Vector Machine (SVM), and Multiple Regression Analysis (MRA). The real production data of a PSS company is used to demonstrate the predictive performance of the proposed model.
According to the experimental results, ANN, SVM, and MRA all have good prediction capabilities of the PSS results. During process adjustment and new product development, the process engineers can utilize our prediction model for process parameter tuning.

中文文獻
陳順宇. (民89). 多變量分析. 台北：華泰書局。
英文文獻
Bui, D. T., Tuan, T. A., Klempe, H., Pradhan, B., & Revhaug, I. (2016). Spatial prediction models for shallow landslide hazards: a comparative assessment of the efficacy of support vector machines, artificial neural networks, kernel logistic regression, and logistic model tree. Landslides, 13(2), 361-378.
Clemen, R. T. (1989). Combining forecasts: A review and annotated bibliography. International Journal of Forecasting, 5(4), 559-583.
Cho, V. (2003). A comparison of three different approaches to tourist arrival forecasting. Tourism Management, 24(3), 323-330.
Chang, Y. C. (2007). Sequencing of chillers by estimating chiller power consumption using artificial neural networks. Building and Environment, 42(1), 180-188.
Chaouachi, A., Kamel, R. M., Andoulsi, R., & Nagasaka, K. (2013). Multiobjective intelligent energy management for a microgrid. IEEE Transactions on Industrial Electronics, 60(4), 1688-1699.
Doumpos, M., & Zopounidis, C. (2009). Monotonic support vector machines for credit risk rating. New Mathematics and Natural Computation, 5(03), 557-570.
Herrera, M., Torgo, L., Izquierdo, J., & Pérez-García, R. (2010). Predictive models for forecasting hourly urban water demand. Journal of Hydrology, 387(1), 141-150.
Hong, W. C., Dong, Y., Zheng, F., & Lai, C. Y. (2011). Forecasting urban traffic flow by SVR with continuous ACO. Applied Mathematical Modelling, 35(3), 1282-1291.
Granger, C. W., & Ramanathan, R. (1984). Improved methods of combining forecasts. Journal of Forecasting, 3(2), 197-204.
Grubb, H., & Mason, A. (2001). Long lead-time forecasting of UK air passengers by Holt–Winters methods with damped trend. International Journal of Forecasting, 17(1), 71-82.
Guyon, I., Weston, J., Barnhill, S., & Vapnik, V. (2002). Gene selection for cancer classification using support vector machines. Machine Learning, 46(1-3), 389-422.
Khan, J., Wei, J. S., Ringner, M., Saal, L. H., Ladanyi, M., Westermann, F., Berthold, F., Schwab, M., Antonescu, C. R., Peterson, C., & Meltzer, P. S. (2001). Classification and diagnostic prediction of cancers using gene expression profiling and artificial neural networks. Nature Medicine, 7(6), 673-679.
Kim, B., Lee, D. W., Park, K. Y., Choi, S. R., & Choi, S. (2004). Prediction of plasma etching using a randomized generalized regression neural network. Vacuum, 76(1), 37-43.
Mohanraj, M., Jayaraj, S., & Muraleedharan, C. (2012). Applications of artificial neural networks for refrigeration, air-conditioning and heat pump systems—A review. Renewable and Sustainable Energy Reviews, 16(2), 1340-1358.
Mubashir, M., Shao, L., & Seed, L. (2013). A survey on fall detection: Principles and approaches. Neurocomputing, 100, 144-152.
Mirowski, P., Chen, S., Ho, T. K., & Yu, C. N. (2014). Demand forecasting in smart grids. Bell Labs Technical Journal, 18(4), 135-158.
Nguyen, H. T., & Nabney, I. T. (2010). Short-term electricity demand and gas price forecasts using wavelet transforms and adaptive models. Energy, 35(9), 3674-3685.
Pradhan, B. (2013). A comparative study on the predictive ability of the decision tree, support vector machine and neuro-fuzzy models in landslide susceptibility mapping using GIS. Computers & Geosciences, 51, 350-365.
Shetty, G. R., Malki, H., & Chellam, S. (2003). Predicting contaminant removal during municipal drinking water nanofiltration using artificial neural networks. Journal of Membrane Science, 212(1), 99-112.
Suykens, J. A., & Vandewalle, J. (1999). Least squares support vector machine classifiers. Neural Processing Letters, 9(3), 293-300.
Taormina, R., & Chau, K. W. (2015). Neural network river forecasting with multi-objective fully informed particle swarm optimization. Journal of Hydroinformatics, 17(1), 99-113.