关键词: 最小二乘支持向量机, 多类分类, 最小输出编码, 阻塞性睡眠呼吸暂停低通气综合征

Abstract: Support Vector Machines（SVM） is one of the most important data mining and machine learning method，and Least Squares Support Vector Machines（LS-SVM） is extended from Support Vector Machines learning algorithm and better in learning speed.The LS-SVM is applied to the multi-class classification methods of SVM including 1-a-r，1-a-1，ECOC and MOC.Then，the multi-class classification methods of LS-SVM are applied to OSAHS diagnoses，which have a good result.

Key words: Least Squares Support Vector Machines, multi-class classification, Minimum Output Coding, Obstructive Sleep Apnea Hypopnea Syndrome