本篇論文主要是透過機器學習(Machine Learning)中支援向量機(Support Vector Machine)演算法對人體心電訊號圖(Electrocardiograph, ECG 以下簡稱心電圖)進行分類。採用穿戴裝置(Wearable device)以便利性與非侵入性檢測(Non-invasive Procedures)的方式來取得人體心電圖，將心電圖數值化後，再將其轉換成數值向量，透過支援向量機演算法找出心電圖之特徵，並對其進行分類。然而，數值化後的心電圖大部分的特徵是不必要的，期望找出必要的特徵後，進而將心電圖作精準的分類。在本研究中，開發出一種逆轉換演算法將心電圖數值化，數值化後經由反覆實驗，發現每一種類別皆有不同的必要特徵，且這些必要特徵能對人體心電圖作出精準的分類。採用非專業醫療儀器，卻成功地對穿戴裝置的心電圖進行分類，達到便利性與非侵入性檢測的目的。

In this thesis, we develop a classifier based on Support Vector Machine for classifying electrocardiograph (ECG) of wearable devices. The wearable device is used to obtain the ECG of the human body in a convenient and non-invasive way. After the ECG is digitized, it is converted into a numerical vector, and the features of the ECG are found and classified through the support vector machine algorithm. However, most of the features of the digitized ECG are unnecessary, and it is expected that ECG can be accurately classified after finding the necessary features. We also develop an inverse transformation algorithm to digitize the ECG in this study. Repeated experiments with digitized ECGs, it was found that each category has different necessary feature which can accurately classify the human ECG. The development of an algorithm to convert the ECG into digitized form and the accurate classification of the ECG are the focus of this paper. Instead of using the professional medical instruments, the purposes of convenience and non-invasive detection are achieved as the ECG measured by the wearable device is successfully classified.

[1] 王培中,透視婦產科：看懂超音波，孕產攻略升級, 山谷文教, 2018.
[2] M. Thaler, The only EKG book you'll ever need, Wolters Kluwer Health, seventh edition, 2012.
[3] 張俐婷, "由2016 CES看智慧穿戴產品發展趨勢," MIC 資訊科技 - 智慧生活科技, 2016.
[4] C. Cortes, V. Vapnik, "Support-vector networks," Maching learning, pp. 273–297,1995.
[5] Q. Firdaus, R. Sigit, T. Harsono and A. Anwar, "Lung cancer detection based on CT-scan images with detection features using gray level co-occurrence matrix (GLCM) and support vector machine (SVM) methods," International Electronics Symposium (IES), pp. 643-648, 2020.
[6] G. Zhang, "A modified SVM classifier based on RS in medical disease prediction," 2009 Second International Symposium on Computational Intelligence and Design, pp. 144-147, 2009.
[7] B. Shanmugapriya and T. Ramakrishnan, "Segmentation of brain tumors in computed tomography images using SVM classifier," International Conference on Electronics and Communication Systems (ICECS), pp. 1-3,2014.
[8] G.Y. Yu, "A finger pulse recognition system based on support vector machines," M.S. thesis, National Taiwan University of Techology and Science, Taipei, Taiwan, 2003.
[9] N.Cristianini, J.Shawe-Taylor, An introduction to support vector machines and other kernel-based learning methods, Cambridge university press, 2000.
[10] C.J. Burges, "A tutorial on support vector machines for pattern recognition," Data Mining and Knowledge Discovery, pp. 955–974, 1998.
[11] A. Buades, B. Coll and J. -. Morel, "Non-Local Means denoising," Image Processing On Line(IPOL), vol. 1 , pp. 208-212, 2011.
[12] A. Buades, B. Coll and J. -. Morel, "A non-local algorithm for image denoising," IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05), vol. 2 , pp. 60-65, 2005.
[13] E. Kreyszig, Advanced Engineering Mathematics, Wiley, fourth edition, 2001.