摘要
大脑通过神经元动作电位即锋电位进行信号的传递、交流和处理,对神经锋电位活动的记录检测是神经科学研究的前提。神经锋电位信号主要是通过细胞外电极进行记录。然而,单个电极上记录到的信号往往是几个相邻神经元的锋电位与大量噪声的叠加。为了从电极记录信号中获得有用的神经元放电信息,对神经元放电序列进行甄别就尤为重要,有必要把每个神经元发出的锋电位从记录信号中分离出来。
模式识别通过数据的先验知识和统计信息来对数据进行分类。本文对模糊聚类和支持向量机等模式识别方法进行深入研究,提出一些简单有效的分类方法解决神经锋电位信号识别中的难点问题。
本文的创新性工作有:
1.实际检测到的锋电位信号往往包含大量噪声和野值点,针对该问题,提出鲁棒模糊聚类方法提高锋电位的分类精度。在聚类过程中,样本点相对于聚类的模糊隶属度不仅与聚类中心的距离有关,还与样本点局部密度值有关。通过减少具有较小密度值的噪声和野值点的模糊隶属度来降低它们在聚类过程中的影响,同时减少聚类间边界点对于各个聚类的隶属度,使聚类更好地分离开,实现对锋电位信号的准确分类。并在此基础上,提出了改进的模糊聚类有效性评价指标,实现在噪声情况下对锋电位数据聚类个数的识别。
2.神经元爆发式放电和记录电极漂移导致单一锋电位聚类形状发生无规则变化,划分聚类方法难以获得满意分类结果。提出基于模糊C均值的层次聚类方法解决这一难题。首先,使用模糊聚类方法将锋电位数据划分为多个关系紧密的小类。然后,通过模糊隶属度来计算次聚类间的连接强度,从整体上衡量划分间的相似性,对小类逐步合并,正确识别复杂的聚类结构。最后,通过样本点与多个中心的平均加权距离来度量类内和类间距离,正确反映出复杂聚类的松紧程度,实现对形状发生无规则变化的锋电位聚类个数地判断。
3.针对锋电位信号的叠加问题,提出基于监督分类的模板匹配方法。根据多类支持向量机的分类特点,设计一种新的训练策略,即在标号为负的训练样本中引入人工合成的叠加锋电位波形,对叠加锋电位信号准确判断,然后通过模板提取对检测到的叠加信号进行分离。对于证据理论神经网络分类器,根据其能识别模糊输入向量的特点,通过设置恰当的阈值实现对叠加信号的识别,在分类过程中对这些叠加信号逐步分离。
4.对支持向量机及其在锋电位分类中的应用进行深入研究。提出一种基于密度信息的加权支持向量机,根据样本点在高维特征空间密度大小调整其与分类平面的距离,减少噪声和野值点的影响,突出具有较大密度值的重要样本点的作用,从而把训练样本的密度分布信息用于构造最优分类平面。然后,提出典型性支持向量机,根据训练样本的分布特点,在高维特征空间选择一些能代表每一类样本分布信息的样本点,在构造最优分类平面过程中,保证训练样本被正确区分的前提下,最大化与典型性样本的距离,把更多恰当的训练样本分布信息用来构造分类平面,提高支持向量机泛化性能。在对仿真神经锋电位数据的实验中,典型性支持向量机取得了较好的分类结果。
The extracellular recording of neural spike activity is a prerequisite for studying information transmission and processing in the brain. The spike recordings are usually obtained with electrodes. However, the recording in single electrode contains spikes from several neurons adjacent to the electrode and a high amount of background noise. Therefore, it is necessary to identify the neural spikes and find out the number of neurons contributing to the electrode recording before further analysis is carried out.
Pattern recognition aims to classify data based on either a priori knowledge or on statistical information extracted from the patterns. In this dissertation, the pattern recognition methods of fuzzy clustering and support vector machine (SVM) are studied deeply, several effective classification methods are proposed to deal with the difficulties in spike sorting.
The main contributions of this dissertation are summarized as follows:
1. A robust fuzzy clustering method is proposed to reduce the influence of noise and outliers in spike sorting. The fuzzy membership degrees are adjusted according to the density values of data points. The noise and outliers with lower density values will have small influence in clustering process. Moreover, the border data between different clusters with low density values will have low memberships to any cluster, which make clusters well separated. In order to obtain the optimal number of clusters with ellipsoidal shapes, an extended fuzzy cluster-validity index is proposed. The fuzzy separation and compactness of clusters are evaluated using the weighted Mahalanobis distances between clusters in the index. The robust fuzzy clustering method is able to classify the real neural spikes with noisy data and outliers.
2. In the presence of spike bursts or electrode drift, the spike waveforms generated by single neuron are varying with time, and then the spike clusters will be smeared and have non-convex shapes. An unsupervised hierarchical clustering based on fuzzy C means is proposed to resolve the problem. The initial clusters are obtained by fuzzy clustering method. Considering the interrelations among initial clusters and the complicated structures of spike clusters, the similarity small clusters are merged based on fuzzy membership degrees. The optimal cluster number is obtained by improved Dunn’s index. The index adopts the weighted distances to calculate within-cluster scatter and between-cluster separation, which is suited for clusters with complicated structures.
3. A template matching based on supervised classification method is proposed to decompose the overlapping spike waveforms. A new training method is designed for multi-class SVM, the overlapping spikes are identified by introducing synthesized overlapping waveforms into training sets. The detected overlapping spike waveforms are decomposed by template extraction. For evidence-theoretic neural network, the overlapping spikes are directly detected by predetermined thresholds, and then they are decomposed in classification process step by step.
4. The SVM and its application in spike sorting are studied deeply. A weighted SVM is proposed based on density information, the density values of training data in feature space is used to adjust the distance from the hyperplane to them. The data with high density will be important to construct the separating hyperplane, while the noisy data with low density will have small influence. And then, the representative SVM is proposed to further improve the classification performance. Some important data are selected according to the distribution characteristic of training data, which should represent the distribution information. The separating hyperplane is constructed by maximizing the distance between the representative vectors and the hyperplane with all the training data being classified correctly. In this way, more useful information of the training data far away from the hyperplane is introduce to reduce the influence of outliers and improve the generalization ability of the learning machine. The proposed methods are applied to the simulated neural spikes and the representative SVM exhibits better classification performance.
引文
[1] Boguski M.S. Bioinformatics. Current Opinion in Genetics and Development. 1994, 4(3): 383-388.
[2] Benton D. Bioinformatics: principles and potential of a new multidisciplinary tool. Trends in Biotechnology. 1996, 14(8): 261-2727.
[3] Jain E. Current trends in bioinformatics. Trends in Biotechnology. 2002,20(8):317-319.
[4]孙啸,陆祖宏,谢建明.生物信息学基础.北京:清华大学出版社. 2005:1-35.
[5] DeCharms R.C., Zador A. Neural representation and the cortical code. Annual Review Neuroscience. 2000, 23: 613-47.
[6] Jenkin M., Harris L. Computational and psychophysical mechanisms of visual coding. New York: Cambridge university press. 1997.
[7] Rieke F., Warland D., Stevenck D.R.V., et al. Spike: exploring the neural code. Cambridge, MA: MIT Press. 1996.
[8] Schmidt E.M. Computer separation of multi-unit neuroelectric data: a review. Journal of Neuroscience Methods. 1984, 12: 95-111.
[9] Lewicki M.S. A review of methods for spike sorting: the detection and classification of neural action potentials. Network: Computation Neural System. 1998, 9: 53-78.
[10]尼克尔斯,马丁,华莱士等,杨雄里译.神经生物学—从神经元到脑.北京:科学出版社,2003: 11-42.
[11] Gersterin G.L., Clark W.A. Simultaneous studies of firing patterns in several neurons. Science. 1964, 143: 1325-1327.
[12] Prochazka V.J., Conrad B., Sindermann F. A neuroelectric signal recognition system. Electroencephalography and Clinical Neurophysiology. 1972, 32: 95-97.
[13] Wilson, M.A., McNaughton, B.L. Dynamics of the hippocampal ensemble code for space. Science. 1993, 261: 1055-1058.
[14] O’Keefe J.O., Reece M.L. Phase relationship between hippocampal place units and the EEG theta rhythm. Hippocampus. 1993, 3: 317-330.
[15] Wood F., Black M.J., Vargas-Irwin C. On the variability of manual spike sorting. IEEE Transactions on Biomedical Engineering. 2004, 51(6): 912-918.
[16] Chen A.H., Zhou Y., Gong H.Q., et al. Chicken retinal ganglion cells response characteristics:multi-channel electrode recording study. Science China (Ser C). 2003, 33: 82-88.
[17] Patrick J.R., Rasmus S.P., Stefano B., et al. Examination of the spatial and temporal distribution of sensory cortical activity using a 100-electrode array. Journal of Neuroscience Methods. 1999, 90(1): 57-66.
[18] Edward W.K., Scott J.N., Nicholas A.J.B., et al. Acute toxicity screening of novel AChE Inhibitors using neuronal networks on microelectrode arrays. NeuroToxicology. 2001, 22(1): 3-12.
[19] Heuschkel M.O., Fejtlb M., Raggenbassc M., et al. A three-dimensional multi-electrode array for multi-site stimulation and recording in acute brain slices. Journal of Neuroscience Methods. 2002, 114(2): 135-148.
[20] Welsh J. P., Schwarz C. Multielectrode recording from the cerebellum. Methods for Neural Ensemble Recordings. 1999, 5: 79–100.
[21] Bankman I.N., Janselewitz S.J. Neural waveform detector for prosthesis control. Proceedings of the 17th IEEE EMBS. 1995, 963-964.
[22] Mukhopadhyay S., Ray G.C. A new interpretation of nonlinear energy operator and its efficacy in spike sorting. IEEE Transactions on Biomedical Engineering. 1998, 45: 180-187.
[23] Choi J.H., Jung H.K., Kim T. A new action potential detector using the MTEO and its effects on spike sorting systems at low signal-to-noise ratios. IEEE Transactions on Biomedical Engineering. 2006, 53(4): 738-746.
[24] Kim K.H., Kim S.J. Neural spike sorting under nearly 0-dB signal-to-noise ratio using nonlinear energy operator and artificial neural-network classifier. IEEE Transactions on Biomedical Engineering. 2000, 47: 1406-1411.
[25] Bankman I.N., Johnson K.O., Schneider W. Optimal detection, classification, and superposition resolution in neural waveform recordings. IEEE Transactions on Biomedical Engineering, 1993, 40(8): 836–841.
[26] Gozani S.N., Miller J.P. Optimal discrimination and classification of neuronal action potential waveforms from multiunit, multichannel recordings using software-based linear filters. IEEE Transactions on Biomedical Engineering. 1994,41(4): 358–372.
[27] Nenadic Z., Burdic J.W. Spike detection using the continuous wavelet transform. IEEE Transactions on Biomedical Engineering. 2005, 52(1): 74-87.
[28] Hulata E., Segev R., Shapira Y., et al. Detection and sorting of neural spikes using wavelet packets. Physical Review Letters. 2000, 85: 4637–4640.
[29] Kim K., Kim S. A wavelet-based method for action potential detection from extracellular neural signal recording with low signal-to-noise ratio. IEEE Transactions on Biomedical Engineering. 2003, 50(8): 999–1011.
[30] Nakatani H., Watanabe T., Hoshimiya N. Detection of nerve action potentials under low signal-to-noise ratio condition. IEEE Transactions on Biomedical Engineering. 2001, 48 (8): 845–849.
[31] Brychta R.J., Tuntrakool S., Appalsamy M., et al. Wavelet methods for spike detection in mouse renal sympathetic nerve activity. IEEE Transactions on Biomedical Engineering. 2007, 54(1): 82-93.
[32] Song M.Z., Wang H.B. A spike sorting framework using nonparametric detection and incremental clustering. Neurocomputing. 2006, 69: 1380-1384.
[33] Roberts W.M. Optimal recognition of neuronal waveforms. Biological Cybernetics. 1979, 35: 73-80
[34] Pouzat C., Mazor O., Lauren G. Using noise signature to optimize spike-sorting and to assess neuronal classification quality. Journal of Neuroscience Methods. 2003, 122: 43-57.
[35] Feldman J.H., Roberge F.A. Computer detection and analysis of neuronal spike sequences. Informatica. 1971, 9: 185-197.
[36] Dinning G.H. Real-time classification of multiunit neural signals using reduced feature sets. IEEE Transactions on Biomedical Engineering. 1981, 28: 804-812.
[37] Wheeler B.C., Heetderks W.J. A comparison of techniques for classification of multiple neural signals. IEEE Transactions on Biomedical Engineering. 1982, 29: 752-759.
[38] Glaser E.M., Marks W.B. On-line separation of interleaved neuronal pulse sequences. Data Acquisition and Processing in Biology and Medicine.1968, 5: 137–56.
[39] Glaser E.M. Separation of neuronal activity by waveform analysis. Advances in Biomedical Engineering. New York: Academic. 1971, 77–136.
[40] Gerstein G.L., Bloom M.J., Espinosa I.E., et al. Design of a laboratory for multineuron studies. IEEE transactions on systems man and cybernetics. 1983, 13: 668–76.
[41] Jung H.K., Choi J.H., Kim T. Solving alignment problems in neural spike sorting using frequency domain PCA. Neurocomputing. 2006, 69: 975-978.
[42] Hulata E., Segev R., Ben-Jacob E. A method for spike sorting and detection based on wavelet packets and Shannon’s mutual information. Journal of Neuroscience Methods. 2002, 117: 1-12.
[43] Letelier J., Weber P. Spike sorting based on discrete wavelet transform coefficients. Journal ofNeuroscience Methods. 2001,101: 93–106.
[44] Oweiss K., Anderson D. Noise reduction in multichannel neural recordings using a new array wavelet denoising algorithm. Neurocomputing. 2001, 38: 1687–1693.
[45] Aksenova T.I., Chibirova O.K., Dryga O.A.,et al. An unsupervised automatic method for sorting neuronal spike waveforms in awake and freely moving animals. Methods. 2003, 30(2): 178-187.
[46] Chibirovaa O.K., Aksenovaa T.I., Benabida A., et al. Unsupervised Spike Sorting of extracellular electrophysiological recording in subthalamic nucleus of Parkinsonian patients. Biosystems. 2005, 79: 159-171.
[47] Gray C.M., Maldonado P.E., Wilson M., et al. Tetrodes markedly improve the reliability and yield of multiple single-unit isolation from multiunit recordings in cat striate cortex. Journal of Neuroscience Methods. 1995, 63: 43-54.
[48] McNaughton B.L., O’Keefe J., Barnes C.A. The stereotrode: a new technique for simultaneous isolation of several single units in the central nervous system from multiple unit records. Journal of Neuroscience Methods. 1983, 8: 391-397.
[49] Recce M.L., O’Keefe J. The tetrode: a new technique for multi-unit extracellular recording. Society for Neuroscience Abstracts. 1989, 15: 1250.
[50] McNaughton T.G., Horch K.W. Action potential classification with dual channel intrafascicular electrodes. IEEE Transactions on Biomedical Engineering. 1994, 41: 609-616.
[51] Capowski J.J. The spike program: a computer system for analysis of neurophysiological action potentials. Computer Technology in Neuroscience (Washington DC: Hemisphere). 1976, 237–51.
[52] Millecchia R., McIntyre R. Automatic nerve impulse identification and separation. Computer Biomedical Research. 1978, 11: 459–68.
[53] Yang X., Shamma S.A. A totally automated system for the detection and classification of neural spikes. IEEE Transactions on Biomedical Engineering. 1988, 35: 806–16.
[54] D’Hollander E.H., Orban G.A. Spike recognition and on-line classification by an unsupervised learning system. IEEE Transactions on Biomedical Engineering. 1979, 26: 279–84.
[55] Jansen R.F., TerMaat A. Automatic waveform classification of extracellular multineuron recordings. Journal of Neuroscience Methods. 1992, 42: 123-132.
[56] Goodall E.V., Horch K.W. Separation of action potentials in multiunit intrafascicular recording. IEEE Transactions on Biomedical Engineering. 1992, 39(3): 289-295.
[57] Stein R.B., Andreassen S., Oguztoreli M.N. Mathematical analysis of optimal multichannel filtering for nerve signals. Biological cybernetics. 1979, 32: 19–24.
[58] Roberts W.M., Hartline D.K. Separation of multiunit nerve impulse trains by a multichannel linear filter algorithm. Brain Research. 1975, 94: 141–149.
[59] Oguztoreli M.N., Stein R.B. Optimal filtering of nerve signals. Biological Cybernetics.1977, 27:41–48.
[60] Zhang P.M., Wu J.Y., Zhou Y., et al. Spike sorting based on automatic template reconstruction with a partial solution to the overlapping problem. Journal of Neuroscience Methods. 2004,135: 55-65.
[61] Lewicki M.S. Bayesian modeling and classification of neural signals. Neural Computation. 1994, 6: 1005-1030.
[62] Atiya A.F. Recognition of multiunit neural signals. IEEE Transactions on Biomedical Engineering. 1992, 39: 723–729.
[63] Chandra R., Optican L.M. Detection, classification, and superposition resolution of action potentials in multiunit single-channel recordings by on-line real-time neural network. IEEE Transactions on Biomedical Engineering. 1997, 44(5): 403-412.
[64] Xu R., Wunsch D. Survey of clustering algorithms. IEEE Transactions on Neural Networks. 2005, 16(3): 645-678.
[65] Jain A.K., Murty N.N. and Flynn P.J. Data clustering: a review. ACM Computing Surveys. 1999, 31(3): 264-323.
[66] Zouridaks G., Tam D.C. Identification of reliable spike templates in multi-unit extracellular recording using fuzzy clustering. Computer Methods and Programs in Biomedicine. 2000, 61: 91-98.
[67] Kaneko H., Suzuki S.S., Okada J., et al. Multineuronal spike classification based on multisite electrode recording, whole-waveform analysis, and hierarchical clustering. IEEE Transactions on Biomedical Engineering. 1999, 46(3): 280-290.
[68] Sahani M.M. The latent variable models for neural data analysis [PhD Thesis]. Pasadena, CA: California Institute of Technology. 1999.
[69] Shoham S., Fellows M.R., Normamn R.A. Robust, automatic spike sorting using mixtures of multivariate t-distributions. Journal of Neuroscience Methods. 2003, 127: 111-122.
[70] Kim K., Kim S. Method for unsupervised classification of multiunit neural signal recording under low signal-to-noise ratio. IEEE Transactions on Biomedical Engineering. 2003, 50(4):421-431.
[71] Nguyen D.P., Frank L., Brown E.N. An application of reversible-jump Markov chain Monte Carlo to spike classification of multi-unit extracellular recordings. Network: Computation in Neural Systems.2003, 14: 61–82.
[72] Mirfakhraei K., Horch K. Classification of action potentials in multiunit intrafascicular recordings using neural network pattern-recognition techniques. IEEE Transactions on Biomedical Engineering. 1994, 41(1): 89-91.
[73] Ohberg F., Johansson H., Bergenheim M., et al. A neural network approach to real-time spike during simultaneous recording from several multi-unit nerve filaments. Journal of Neuroscience Methods. 1996, 64: 181-187.
[74] Hermle T., Schwarz C., Bogdan M. Employing ICA and SOM for spike sorting of multielectrode recordings from CNS. Journal of Physiology– Paris. 2004, 98: 349-356.
[75] Delesclue M., Pouzat C. Efficient spike-sorting of multi-state neurons using inter-spike intervals information. Journal of Neuroscience Methods. 2006, 150: 16-29.
[76] Pouzat C. Technique for spike-sorting. Methods and models in neurophysics. Berlin:Springer-Verlag. 2005: 729–786.
[77] Pouzat C., Delescluse M., Viot P., et al. Improved spike-sorting by modeling firing statistics and burst-dependent spike amplitude attenuation: a Markov chain Monte Carlo approach. Journal of Neurophysiology. 2004, 91: 2910–2928.
[78] Harris K.D., Henze D. A., Csicsvari J. Accuracy of tetrode spike separation as determined by simultaneous intracellular and extracellular measurements. Journal of Neurophysiology. 2000, 84: 401-414.
[79] Fee M.S., Mitra P.P., Kleinfeld D. Automatic sorting of multiple-unit neuronal signals in the presence of anisotropic and non-Gaussian variability. Journal of Neuroscience Methods. 1996a, 69: 175–88.
[80] Fee M.S., Mitra P.P., Kleinfeld D. Variability of extracellular spike waveforms of cortical-neurons. Journal of Neurophysiology. 1996b, 76: 3823–33.
[81] Snider R.K., Bonds A.B. Classification of non-stationary neural signals. Journal of Neuroscience Methods. 1998, 84 (1): 155-166.
[82] Emondi A.A., Rebrik S.P., Kurgansky A.V., et al. Tracking neurons recorded from tetrodes across time. Journal of Neuroscience Methods. 2004, 135 (1-2): 95-105.
[83] Bar-Hillel A., Spiro A., Stark E. Spike sorting: Bayesian clustering of nonstationary data.Proceedings of the 18th International Conference on Neural Information Processing Systems. 2004, 17: 105–112.
[84] Bar-Hillel A., Spiro A., Stark E. Spike sorting: Bayesian clustering of non-stationary data. Journal of Neuroscience Methods. 2006, 157: 303-316.
[85] Takahashi S., Anzai Y., Sakurai Y. Automatic sorting for multi-neuronal activity recorded with tetrodes in the presence of overlapping spikes. Journal of Neurophysiology. 2003, 89: 2245-2258.
[86] Takahashi S., Sakurai Y., Tsukada M., et al. Classification of neuronal activities from tetrode recordings using independent component analysis. Neurocomputing. 2002, 49: 289–298.
[87] Takahashi S., Sakurai Y. Real-time and automatic sorting of multi-neuronal activity for sub-millisecond interactions in vivo. Neuroscience. 2005, 134: 301–315.
[88] Mamlouk A.M., Sharp H., Menne K.M.L. Unsupervised spike sorting with ICA and its evaluation using GENESIS simulations. Neurocomputing. 2005, 65: 275-282.
[89] Bezdek J.C. Pattern recognition with fuzzy objective function algorithms. New York, Plenum. 1981.
[90] Bezdek J.C., Pal S.K. Fuzzy models for pattern recognition-methods that search for structures in data. Piscataway, NJ: IEEE Press. 1992.
[91] Krishnapuram R., Keller J. A possibilistic approach to clustering. IEEE Transaction on Fuzzy Systems. 1993, 1(2): 98-110.
[92] Krishnapuram R., Keller J. The possibilistic c-menas algorithm: insights and recommendations. IEEE Transaction on Fuzzy Systems. 1996; 4(3): 385-393.
[93]边肇祺,张学工.模式识别.北京:清华大学出版社. 2000: 280-282.
[94] Frigui H., Krishnapuram R. A robust algorithm for automatic extraction of an unknown number of clusters from noisy data. Pattern Recognition Letters. 1996, 17: 1223-1232.
[95]刘小芳,曾黄麟,吕炳朝.点密度函数加权模糊c-均值算法的聚类分析.计算机工程与应用. 2004, 24: 64-65.
[96] Rosenblatt M. Remarks on some nonparametric estimates of a density function. Annals of Mathematical Statistics. 1956, 27 (6): 832-837.
[97] Parzen E. On estimation of a probability density function and mode. Annals of Mathematical Statistics. 1962 , 33 (8): 1065-1076.
[98]王星.非参数统计.北京:中国人民大学出版社. 2005: 213-218.
[99] Rudemo M. Empirical choice of histograms and kernel density estimation. Scandinavian Journalof Statistics. 1982, (9): 65-78.
[100] Hall P., Marron J.S., Park B.V. Smoothed cross validation. Probability Theory and Related Field. 1992, 90: 149-173.
[101] Pakhiram M.K., Bandyopadhyay S., Maulik U. Validity index for crisp and fuzzy clustering. Pattern Recognition. 2004, 37(3): 487-501.
[102] Xie L.X., Beni G. A Validity Measure for Fuzzy Clustering. IEEE Transactions on Pattern Analysis and Machine Intelligence. 1991, 13(8): 841-847.
[103] Bouguessa M., Wang S., Sun H. An objective approach to cluster validation. Pattern Recognition Letters. 2006, 27: 1419-1430.
[104] Bezdek J.C., Pal N.R. Some new indexes of cluster validity. IEEE Transactions on System, Man, and Cybernetics-Part B: Cybernetics. 1998, 28 (3): 301-315.
[105] Wang G.L., Zhou Y., Chen A.H., et al. A robust method for spike sorting with automatic overlap decomposition. IEEE Transactions on Biomedical Engineering. 2006, 53: 1195-1198.
[106] Vapnik V. Statistical Learning Theory. New York: Wiley, 1998.
[107] Cristianini N., Shawe-Taylor J. An Introduction to Support Vector Machines. Cambridge, UK: Cambridge University Press. 2000.
[108] Sch?lkopf P. and Smola A.J. Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond. Cambridge, MA: MIT Press. 2002.
[109] Sch?lkopf B., Smola A.J., Willianson R. New support vector algorithms. Neural Computation. 2000, 12 (5): 1207–1245.
[110] Burges C.J.C. A tutorial on support vector machines for pattern recognition. Data Mining Knowledge Discovery. 1998, 2(2): 1–47.
[111] Hsu C.W., Lin C.J. A comparison on methods for multi-class support vector machines. IEEE Transactions on Neural Networks. 2002, 13: 415-425.
[112] Denoeux T. A neural network classifier based on Dempster-Shafer theory. IEEE Transaction on Systems, Man, and Cybernetics-part A: Systems and Humans. 2000, 30(2): 131-150.
[113] Lee K.Y., Kim D.W., Lee K.H. Possibilistic support vector machines. Pattern Recognition. 2005, 38: 1325-1327.
[114] Tao Q., Wu G. W., Wang F. Y. Posterior probability support vector machines for unbalanced data. IEEE Transactions on Neural Networks. 2005, 6: 1561-1573.
[115] Zhang X.G. Using class-center vectors to build support vector machines. NNSP'99, 1999.
[116] Kou Z., Xu J., Zhang X.G., et al. An improved support vector machine using class-medianvectors. Proceedings of 8th international conference on neural information processing, Shanghai, China. 2001, 2: 883-887.
[117] Platt J.C. Fast training of support vector machines using the sequential minimal optimization. Advances in Kernel Methods: Support Vector Machines. Cambridge, MA: MIT Press.