本篇論文研究的目的是透過腦電波應用多類別支持向量機（SVM）對睡眠階段來進行分類。這裡將睡眠階段劃分為清醒、快速眼動期、以及三個非快速動眼期（N1, N2, N3）共五個階段。在我們的研究中，腦電波的特徵是由散射變換所擷取，再將提取後的特徵利用多類別支持向量機進行睡眠階段的預測。我們也使用典型相關分析（CCA）對資料進行降維並融合兩個腦電圖（EEG）信號的信息。這裡我們利用留一法則交叉驗證（LOSOCV）來評估所提出算法的效能，並將結果與專家判讀進行比較。此外，資料類別不平衡是分類中的常見問題，這裡使用合成少數過採樣技術（SMOTE）用以平衡數據。然而，有時在數據中會存在一些異常點，因此我們提出了應用於多類別支持向量機上之敏感度曲線來檢測可能影響分類結果的資料。最後我們在研究中證明了支持向量機通過腦電圖信號自動識別睡眠階段的能力。

The objective of this study is to identify sleep stages divided into Awake, REM, N1, N2, and N3 by applying a multi-class support vector machine (SVM) on brain waves. In this work, the features of the brain waves are first extracted by the scattering transform, then are used to train multi-class SVM for sleep stage prediction. We also use canonical correlation analysis (CCA) for dimension reduction and fused the information of two Electroencephalography (EEG) signals. We consider leave-onesubject-out cross-validation (LOSOCV) to evaluate the performance of the proposed algorithm. And the results are compared with human expert classification. A class imbalance is a common problem in data classification, then a Synthesized Minority Oversampling Technique (SMOTE) is used to balance data here. However, sometimes there are some abnormal points that may affect the classification results in the data. Hence we propose the sensitivity curve to detect these abnormal points. The simulated example and two real data demonstrate the capability of using the SVM combine with (or integrate) CCA on the EEG signals for automatically recognizing sleep stages. Furthermore, the SC detect outliers successfully then improve the prediction accuracy for sleep stages based on the SVM.

[1] J. Anden and S. Mallat. Deep scattering spectrum IEEE Transactions on Signal Processing, 62(16):4114-4128, 2014.
[2] Berry, D.G. Budhiraja, et al. Rules for scoring respiratory events in sleep:update of the 2007 AASM manual for the scoring of sleep and associated events J. Clin. Sleep Med., 8(5):597–619, 2012.
[3] Nicolle M. Correa, Tulay Adali, Yi-Ou Li, Vince D. Calhoun Canonical Correlation Analysis for Data Fusion and Group Inferences. IEEE Signal Processing Magazine, 62(16):4114-4128, 2010.
[4] Nitesh V. Chawla, Kevin W. Bowyer, Lawrence O. Hall, and W. Philip Kegelmeyer. Smote: Synthetic minority over-sampling technique. J. Artif. Int. Res., 27(4):39 - 50, 2002
[5] A. Christmann, I. Steinwart. On Robustness Properties of Convex Risk Minimization Methods for Pattern Recognition. Journal of Machine Learning Research 5, 1007–1034, 2004.
[6] K. Crammer and Y. Singer On the Algorithmic Implementation of Multiclass Kernel-based Vector Machines Journal of Machine Learning Research, 2:265-292, 2001.
[7] N. Cristianini and J. Shawe-Taylor. An introduction to Support Vector Machines and Other Kernel-based Learning Methods Cambridge: Cambridge University Press., 2000.
[8] Frank R. Hampel The influence curve and its role in robust estimation. Journal of the American Statistical Association, 69(346):383-393, 1974.
[9] Frank R. Hampel, Elvezio M. Ronchetti, Peter J. Rousseeuw, and Werner A. Stahel. Robust Statistics: the approach based on influence function, 1986.
[10] C. Iber, S. Ancoli-Isreal, A. L. Chesson Jr., S. Quan The AASM Manual for the Scoring of Sleep and Associated Events - Rules Terminology and Technical Specification, 2007.
[11] H. Hotelling Relations Between Two Sets of Variates Biometrika, 28(3):321-377, 1936
[12] Gi-Ren Liu, Yu-Lun Lo, John Malik, Yuan-Chung Sheu, and Hau-Tieng Wu. Diffuse to fuse EEG spectra – Intrinsic geometry of sleep dynamics for classification. Biomedical Signal Processing and Control, 55:101576, 2020.
[13] Gi-Ren Liu, Caroline Lustenberger, Yu-Lun Lo, Wen-Te Liu, Yuan-Chung Sheu, and Hau Tieng Wu. Save Muscle Information-Unfiltered EEG Signal Helps Distinguish Sleep Stage. Sensors, 20(7), 2020.
[14] J. Mercer. Functions of positive and negative type, and their connection with the theory of integral equations. Proceedings of the Royal Society of London, Series A: Containing Paper of a Mathematical and Physical Character, 209:415-446, 1909.
[15] Rechtschaffen, A. Kales A Manual of Standardized Terminology, Techniquesand Scoring System for Sleep Stages of Human Subjects. US Government Printing Office, Washington, 1968.
[16] B. Scholkopf and Alexander J. Smola Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond. MIT Press, 2001.
[17] John W. Tukey. A survey of sampling from contaminated distributions, 1959.
[18] John W. Tukey. Exploratory Data Analysis Reading, Mass: Addison-Wesley Pub. Co.,1977.
[19] Vladimir N. Vapnik. Statistical Learning Theory. Wiley-Interscience, 1998.
[20] J. Weston and C. Watkins. Support vector machines for multi-class pattern recognition. In Proceedings of the Seventh European Symposium on Artificial Neural Networks, 1999.