Character-Recognition(Lib-SVM)

  支持向量机的研究现已成为机器学习领域中的研究热点，其理论基础是Vapnik[3]等提出的统计学习理论。统计学习理论采用结构风险最小化准则，在最小化样本点误差的同时，缩小模型泛化误差的上界，即最小化模型的结构风险，从而提高了模型的泛化能力，这一优点在小样本学习中更为突出。SVM理论正是在这一基础上发展而来的，经过十几年的研究和发展，已开始逐步应用于一些领域。在解决小样本、非线性及高维模式识别问题中表现出许多特有的优势，已经在模式识别、函数逼近和概率密度估计等方面取得了良好的效果。( Support Vector Machine (SVM) is a new machine learning technique in recent years developed based on statistical learning theory (SLT). It wins popularity due to many attractive features and emphatically performance in the fields of nonlinear and high dimensional pattern recognition. The theory and algorithm of SVC is studied at first, then, simulation is to recognize handwritten numeral with the Lib-SVM toolbox. At last, we study the result, which shows that the SVC can do the classification problem with good performance, shorter operation time and is more suitable for real-time implementation.)