摘要
手写数字的识别是模式识别及机器学习的一个重要应用,应用范围非常广泛。本文提出一种基于决策树算法的手写数字识别方法,该方法通过提取基于密度的特征,通过训练得到一个决策树分类模型,进而进行手写数字的识别。实验证明该方法能够快速有效的进行手写数字的识别。
Handwritten numeral recognition is an application of pattern recognition and machine learning and has wide use in many fields. This paper proposed a method based on decision tree. The method can get a decision tree model for identification of handwritten numeral by extracting features based on： density. The experiments show that the proposed method can quickly and effectively recognize handwritten numeral.
出处
《信息技术与信息化》
2011年第6期21-23,73,共4页
Information Technology and Informatization
关键词
手写数字识别
特征提取
模式识别
决策树
Handwritten numeral recognition Feature extraction Pattern recognition Decision tree
