A Kernel-Based Predictive Model for Guillain-Barré Syndrome

详细信息    查看全文

• 关键词：SVM kernels ; Classification ; Performance evaluation ; AUC
• 刊名：Lecture Notes in Computer Science
• 出版年：2015
• 出版时间：2015
• 年：2015
• 卷：9414
• 期：1
• 页码：270-281
• 全文大小：249 KB
• 参考文献：1.Hornik, K., Karatzoglou, A., Smola, A., Zeileis, A.: kernlab - an S4 package for kernel methods in R. J. Stat. Softw. 11(9), 1–20 (2004)
  2.Canul-Reich, J., Hernández-Torruco, J., Frausto-Solis, J., Méndez-Castillo, J.J.: Finding relevant features for identifying subtypes of guillain-barré syndrome using quenching simulated annealing and partitions around medoids (in review)
  3.Cohen, J.: A coefficient of agreement for nominal scales. Educat. Psychol. Meas. 20(1), 37–46 (1960)CrossRef
  4.Chang, C.C., Hsu, C.W., Lin, C.J.: A practical guide to support vector classification. Technical report, Department of Computer Science, National Taiwan University (2003)
  5.Lin, C.J., Hsu, C.W.: A comparison of methods for multiclass support vector machines. IEEE Trans. Neural Netw. 13, 1045–1052 (2002)CrossRef
  6.Fawcett, T.: An introduction to roc analysis. Pattern Recogn. Lett. 27(8), 861–874 (2006)CrossRef MathSciNet
  7.Hand, D.J., Hill, R.J.: A simple generalisation of the area under the roc curve for multiple class classification problems. Mach. Learn. 45(2), 171–186 (2001)CrossRef MATH
  8.Kuwabara, S.: Guillain-barré syndrome. Drugs 64(6), 597–610 (2004)CrossRef MathSciNet
  9.Matsumoto, M., Nishimura, T.: Mersenne twister: a 623-dimensionally equidistributed uniform pseudo-random number generator. ACM Trans. Model. Comput. Simul. 8(1), 3–30 (1998)CrossRef MATH
  10.Meyer, D., Dimitriadou, E., Hornik, K., Weingessel, A., Leisch, F.: e1071: Misc Functions of the Department of Statistics (e1071), TU Wien (2014). R package version 1.6-3
  11.Pascual, S.I.P.: Protocolos Diagnóstico Terapéuticos de la AEP: Neurología Pediátrica. Síndrome de Guillain-Barré. Asociación Espanola de Pediatría, Madrid (2008)
  12.Solokova, M., Lapalme, G.: A systematic analysis of performance measures for classification tasks. Inf. Process. Manage. 45, 427–437 (2009)CrossRef
  13.Uncini, A., Kuwabara, S.: Electrodiagnostic criteria for guillain-barré syndrome: a critical revision and the need for an update. Clin. Neurophysiol. 123(8), 1487–1495 (2012)CrossRef
  14.Vapnik, V.: Statistical Learning Theory. Wiley, New York (1998)MATH
  
• 作者单位：José Hernández-Torruco (16)
  Juana Canul-Reich (16)
  Juan Frausto-Solis (17)
  Juan José Méndez-Castillo (18)
  
  16. División Académica de Informática y Sistemas, Universidad Juárez Autónoma de Tabasco, Cunduacan, Tabasco, Mexico
  17. Instituto Tecnológico de Ciudad Madero, Av. 1o. de Mayo esq. Sor Juana Inés de la Cruz s/n, Col. Los Mangos, 89440, Ciudad Madero, Tamaulipas, Mexico
  18. Hospital General de Especialidades Dr. Javier Buenfil Osorio, Av. Lázaro Cárdenas 208, Col. Las Flores, 24097, San Francisco De Campeche, Campeche, Mexico
  
• 丛书名：Advances in Artificial Intelligence and Its Applications
• ISBN：978-3-319-27101-9
• 刊物类别：Computer Science
• 刊物主题：Artificial Intelligence and Robotics
  Computer Communication Networks
  Software Engineering
  Data Encryption
  Database Management
  Computation by Abstract Devices
  Algorithm Analysis and Problem Complexity
  
• 出版者：Springer Berlin / Heidelberg
• ISSN：1611-3349

文摘

The severity of Guillain-Barré Syndrome (GBS) varies among subtypes, which can be mainly Acute Inflammatory Demyelinating Polyneuropathy (AIDP), Acute Motor Axonal Neuropathy (AMAN), Acute Motor Sensory Axonal Neuropathy (AMSAN) and Miller-Fisher Syndrome (MF). In this study, we use a real dataset that contains clinical, serological, and nerve conduction tests data obtained from 129 GBS patients. We apply Support Vector Machines (SVM) using four different kernels: linear, Gaussian, polynomial and Laplacian to predict four GBS subtypes. We compare SVM results with those of C4.5. We evaluated performance under both 10-FCV and train-test scenarios. Experimental results showed performance of both classifiers are comparable. SVM slightly outperformed C4.5 with Polynomial kernel in 10-FCV. And it did with Laplacian, polynomial and Gaussian kernels in train-test. This is an ongoing research project and further experiments are being conducted.