近年來，冠心病為最為常見的心血管疾病之一，位居心臟病類型的前幾大死因。過往常用的醫學技術如冠狀動脈造影技術，該技術可直接評估冠狀動脈內狹窄或阻塞的位置，然而此檢測技術除了需要耗費昂貴的醫療成本，且常於有相關症狀出現時才會使用此類型的侵入式檢測方法，對於疾病即早治療的幫助十分有限。現今醫療逐漸朝智慧化與數位化發展，並期望改善過往侵入性診斷的高成本與費時性的問題，本研究結合冠心病醫療數據以及人工智慧的方法，期望以數據分析的方式，結合分類模型與超參數架構優化，以有效對冠心病進行高準確率的分類預測。
本研究以深度學習模型為基礎，對冠心病的臨床診斷資料與數據集進行訓練和測試，並結合多種超參數調優演算法，包括隨機搜索法（Random Search）、Tree Parzen Estimator（TPE）和協方差自適應進化策略（Covariance Matrix Adaptation Evolution Strategy，CMA-ES），以優化模型的性能。透過模型改善，我們評估不同超參數調優方法的效果，並比較它們在冠心病分類中的性能。研究結果顯示，透過適當的超參數調優，深度學習模型在冠心病分類中取得了顯著的改善，並呈現出優於傳統方法的效果。

Recent years have seen a rise in coronary heart disease, a prevalent cardiovascular condition and a leading cause of heart-related fatalities. Traditional methods like coronary angiography, while effective in pinpointing artery blockages, are costly and typically utilized only after symptoms emerge, hampering early intervention. To counteract these limitations and align with the trend towards intelligent healthcare, this study merges coronary heart disease data with artificial intelligence techniques. By refining classification models through data analysis and parameter adjustments, the aim of this study is to deliver precise predictions for coronary heart disease classification, thus streamlining diagnosis and management processes.
This study uses deep learning models to analyze clinical diagnostic data for coronary heart disease. We employ different hyperparameter optimization techniques like Random Search, Tree Parzen Estimator (TPE), and Covariance Matrix Adaptation Evolution Strategy (CMA-ES) to enhance model performance. Implementation results show that our proposed approaches using 1D CNN + TPE, 1D CNN + CMA-ES, BPNN + TPE and BPNN + CMA-ES get substantial improvement in coronary heart disease classification compared to traditional approaches

摘要 I
Abstract II
目錄 III
圖目錄 V
表目錄 VI
第一章　緒論 1
1.1研究背景與動機 1
1.2研究目的 2
1.3論文架構 3
第二章　文獻回顧 5
2.1 冠狀動脈心臟疾病 5
2.1.1 冠狀動脈心臟疾病相關 5
2.1.2 檢測方法 6
2.2 深度學習模型 7
2.2.1 倒傳遞神經網路 8
2.2.2 卷積神經網路 9
2.3 超參數調優方法 10
第三章　研究方法 13
3.1 研究架構 13
3.2 資料前處理 16
3.3 深度學習模型之建構 17
3.3.1 倒傳遞神經網路之建構 17
3.3.2 一維卷積神經網路之建構 18
3.4 深度學習模型之超參數最佳化 19
3.4.1 超參數設置 19
3.4.2 隨機搜索法於超參數最佳化 21
3.4.3 Tree Parzen Estimator於超參數最佳化 24
3.4.4 CMA-ES於超參數最佳化 27
第四章　案例研究 30
4.1 案例背景與說明 30
4.2 資料描述 31
4.3 資料前處理 33
4.4 深度學習模型初始建構 34
4.4.1 倒傳遞神經網路初始模型 34
4.4.2 一維卷積神經網路初始模型 36
4.5深度學習模型結合超參數優化演算法於模型之最佳化 38
4.5.1 超參數優化之目標函數與限制範圍 38
4.5.2 深度學習模型結合隨機搜索法於模型之最佳化 40
4.5.3 深度學習模型結合Tree Parzen Estimator於模型之最佳化 44
4.5.4 深度學習模型結合CMA-ES於模型之最佳化 48
4.6實驗結果與討論 52
第五章　結論與未來展望 57
5.1 結論 57
5.2 未來研究方向 58
參考文獻 59

[1] Ambrose, A. J. & Najafi, A. (2018). Strategies for the Prevention of Coronary Artery Disease Complications: Can We Do Better. The American Journal of Medicine, 131(9), 1003-1009.
[2] Ayshwarya, B., Dhanamalar, M., & Sasikumar, V. R. (2023). Heart Diseases Prediction Using Back Propagation Neural Network with Butterfly Optimization. 2023 Fifth International Conference on Electrical, Computer and Communication Technologies (ICECCT), 01-06.
[3] Bergstra, J., & Bengio, Y. (2012). Random Search for Hyper-Parameter Optimization. Journal of Medicine Research, 13, 281-305.
[4] Bergstra, J., Komer, B., Eliasmith, C., Yamins, D., & Cox, D. (2015). Hyperopt: A Python Library for Optimizing the Hyperparameters of Machine Learning Algorithms. Computational Science & Discovery, 8(1), 01-25.
[5] Fazlali, H. R., Karimi, N., Soroushmehr, S. M., Samavi, S., Nallamothu, B., Derksen, H., & Najarian, K. (2015). Robust catheter identification and tracking in X-ray angiographic sequences. Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference, 7901–7904.
[6] Feurer, M., Klein, A., Eggensperger, K., Springenberg, J.T., Blum, M., & Hutter, F. (2015). Efficient and Robust Automated Machine Learning. Advances in Neural Information Processing Systems, 28, 2962-2970.
[7] Hansen, N., & Ostermeier, A. (2001). Completely derandomized self-adaptation in evolution strategies. Evolutionary Computation, 9(2), 159-195.
[8] Hernandez-Vela, A., Gatta, C., Escalera, S., Igual, L., Martin-Yuste, V., Sabate, M., & Radeva, P. (2012). Accurate coronary centerline extraction, caliber estimation and catheter detection in angiographies. IEEE Transactions on Information Technology in Biomedicine, 16(6), 1332-1340.
[9] Hornik, K., Stinchcombe, M., & White, H. (1999). Universal approximation of an unknown mapping and its derivatives using multilayer feedforward networks. Networks, 3(5), 551-560.
[10] Ihsan, M.F., Mandala, S., & Pramudyo, M. (2022). Study of Feature Extraction Algorithms on Photoplethysmography (PPG) Signals to Detect Coronary Heart Disease. 2022 International Conference on Data Science and Its Applications (ICoDSA), 300-304.
[11] Krizhevsky, A., Sutskever, I., & Hinton, G. (2012). ImageNet Classification with Deep Convolutional Neural Networks. Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 25(2), 1106-1114.
[12] Kiranyaz, S., Ince, T., Hamila, R. & Gabbouj, M. (2015). Convolutional Neural Networks for Patient-Specific ECG Classification. Neural Information Processing Systems, 2608–2611.
[13] LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436-444.
[14] LeCun, Y., Bottou L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11), 2278-2324.
[15] Nomura, M., Watanabe., S., Akimoto, Y., Ozaki, Y., & Onishi, M. (2020). Warm Starting CMA-ES for Hyperparameter Optimization. Proceedings of the AAAI Conference on Artificial, 35(10), 9188-9196.
[16] Rohan, B., Ramu, V., Anirban, D. C., & Arpan, P. (2016). Identifying coronary artery disease from photoplethysmogram. Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing: Adjunct, 1084-1088.
[17] Roohallah, A., Mohammad, J. H., Abbas, K., Fahime, K., Mohamad, R., Nizal, S., & Saeid, P. (2018). Non-invasive detection of coronary artery disease in high-risk patients based on the stenosis prediction of separate coronary arteries. Journal of Ambient Intelligence and Humanized Computing, 162(1), 119-127.
[18] Run-jing, Z., & Keyang, L. (2011). Fisher classifier in diagnosis of coronary heart disease. 2011 4th International Congress on Image and Signal Processing, 5, 2654-2658.
[19] Sowmiya, C., & Sumitra, P. (2020). A hybrid approach for mortality prediction for heart patients using ACO-HKNN. Journal of Ambient Intelligence and Humanized Computing, 12(5), 5405-5412.
[20] Taoying, L., Miao, H., & Xu, W. (2020). A Hybrid CNN-LSTM Model for Forecasting Particulate Matter (PM2.5). IEEE Access, 8, 26933-26940.

[21] Thenmozhi, K., & Deepika, P. (2014). Heart Disease Prediction Using Classification with Different Decision Tree Techniques. International Journal of Engineering Research and General Science, 2(6), 6-1.
[22] Yixin, C., Jianye, Z., Yongsheng, L., Siwei, Z., & Jing, C. (2020). Research on the Prediction Method of Ultimate Bearing Capacity of PBL Based on IAGA-BPNN Algorithm. IEEE Access, 8, 179141-179155.
[23] Zhang, H. C., & Huang, S. H. (1955). Applications of neural networks in manufacturing: a state-of-the-art survey. International Journal of Production Research, 33(3), 705-728.
[24] Zhenyu, C., Lutao, W., Bo, Li., Siyu, C., Jingchen, B., Fei, Z., & Yanhong, D. (2022). An Improved Faster R-CNN Transmission Line Bolt Defect Detection Method. 2022 World Automation Congress (WAC), 82-85.
[25] 黃瑞仁（2008）。冠狀動脈心臟病的診斷與治療。中華民國內膜異位症婦女協會會刊，15(2)，7-14。
[26] 衛生福利部國民健康署：2020年全國十大死因統計結果
[27] 開源資料集: Predicting Heart Disease Using Clinical Variables (kaggle.com)